WO2014035989A1 - Flagellin fusion proteins and methods of use - Google Patents

Abstract

Compositions that include a fusion protein comprising flagellin and at least one antigen that has an isoelectric point greater than about 7.0 and that is fused to at least one domain 3 of the flagellin activate Toll-like Receptor 5. Methods of stimulating an immune response, in particular, a protective immune response include administering a composition that includes an antigen fused to a loop of domain 3 of flagellin.

Description

FLAGELLIN FUSION PROTEINS AND METHODS OF USE

RELATED APPLICATION

[0001] This application is a continuation-in-part application of U.S. Application No. 13/931,028, filed on June 28, 2013, which claims the benefit of U.S. Provisional Application No. 61/743,165, filed on August 28, 2012. The entire teachings of the above applications are incorporated herein by reference.

INCORPORATION BY REFERENCE OF MATERIAL IN ASCII TEXT FILE

[0002] This application incorporates by reference the Sequence Listing contained in the following ASCII text file being submitted concurrently herewith: a) File name: 37101054002SEQUENCELISTiNG.txt; created August 26, 2013, 1,174 KB in size.

GOVERNMENT SUPPORT

[0003] This invention was made with government support under

HHSOl 002011000011C from the Biomedical Advanced Research and Development

Authority. The government has certain rights in the invention.

BACKGROUND OF THE INVENTION

[0004] Compositions that include antigens can be employed to stimulate immunity to disease consequent to or infection from exposure to an organism that includes the antigen. Antigens employed in compositions to stimulate immunity, specifically protective immunity to a particular disease, often resemble the disease- causing organism from which the antigen is a source. Compositions that include antigens may also include adjuvants that augment the immune response to the antigen to maximize the production of antibodies that neutralize the related antigen in a disease-causing organism. Recombinant DNA technology has been employed to generate fusion proteins that include antigens and Toll-like Receptor agonists, specifically, flagellin, which is a Toll-like Receptor 5 agonist, to augment an

₁₄98944.1 immune response to the antigen. However, not all antigens are suitable for existing formats of fusion proteins to flagellin. Therefore, a need exists for new and improved designs of fusion protein that include flagellin and an antigen for use in methods of stimulating an immune response, specifically a protective immune response to the antigen.

SUMMARY OF THE INVENTION

[0005] The present invention generally relates to compositions that include fusion proteins that stimulate an immune response, specifically a protective immune response in a subject, and methods of using the fusion proteins.

[0006] In an embodiment, the invention is a composition that includes a fusion protein comprising a flagellin and at least one antigen that has an isoelectric point greater than about 7.0 and is fused to at least one loop of domain 3 of the flagellin. The fusion protein of the invention activates a Toll-like Receptor 5.

[0007] In another embodiment, the invention is a composition that includes a fusion protein comprising a flagellin and at least one antigen that has an isoelectric point greater than about 7.5 and is fused to at least one loop of domain 3 of the flagellin. The fusion protein of the invention activates a Toll-like Receptor 5.

[0008] In another embodiment, the invention is a composition comprising at least three fusion proteins that each activate a Toll-like Receptor 5, wherein: (a) a first fusion protein includes a first flagellin and a first influenza A viral

hemagglutinin antigen that has an isoelectric point greater than about 6.0 fused to a portion of the first flagellin; (b) a second fusion protein includes a second flagellin and a second influenza A viral hemagglutinin antigen that is distinct from the first influenza A viral hemagglutinin antigen and that has an isoelectric point greater than about 7.0 fused to a portion of the second flagellin; and (c) a third fusion protein includes a third flagellin and a first influenza B viral hemagglutinin antigen that has an isolectric point greater than about 8.0 and is fused to at least one loop of domain 3 of the third flagellin.

[0009] In yet another embodiment, the invention is a composition comprising at least four fusion proteins that each activate a Toll-like Receptor 5, wherein: (a) a first fusion protein includes a first flagellin and a first influenza A viral

hemagglutinin antigen that has an isoelectric point greater than about 6.0 fused to a portion of the first flagellin; (b) a second fusion protein includes a second flagellin and a second influenza A viral hemagglutinin antigen that is distinct from the first influenza A viral hemagglutinin antigen and that has an isoelectric point greater than about 7.0 fused to a portion of the second flagellin; (c) a third fusion protein includes a third flagellin and a first influenza B viral hemagglutinin antigen that has an isolectric point greater than about 8.0 and is fused to at least one loop of domain 3 of the third flagellin; and (d) a fourth fusion protein that activates a Toll-like Receptor 5 and includes a second influenza B viral hemagglutinin antigen that is distinct from the first influenza B viral hemagglutinin antigen and that has an isoelectric point greater than about 8.0 fused to at least one loop of domain 3 in the fourth flagellin.

[0010] In another embodiment, the invention is a method of stimulating an immune response to an antigen in a subject, comprising the step of administering to the subject a composition that includes a fusion protein that activates a Toll-like Receptor 5, the fusion protein comprising a flagellin and at least one antigen that has an isoelectric point greater than about 7.0 and that is fused to at least one loop of domain 3 of the flagellin.

[0011] In still another embodiment, the invention is a method of stimulating an immune response to an antigen in a subject, comprising the step of administering to the subject a composition that includes a fusion protein that activates a Toll-like Receptor 5, the fusion protein comprising a flagellin and at least one antigen that has an isoelectric point greater than about 7.5 and that is fused to at least one loop of domain 3 of the flagellin.

[0012] In yet another embodiment, the invention is a method of stimulating an immune response in a subject, comprising the step of administering to the subject a composition that includes at least three fusion proteins each of which activates a Toll-like Receptor 5, wherein: (a) a first fusion protein includes a first flagellin and a first influenza A viral hemagglutinin antigen that has an isoelectric point greater than about 6.0 fused to a portion of the first flagellin; (b) a second fusion protein includes a second flagellin and a second influenza A viral hemagglutinin antigen that is distinct from the first influenza A viral hemagglutinin antigen and that has an isoelectric point greater than about 7.0 fused to a portion of the second flagellin; and (c) a third fusion protein includes a third flagellin and a first influenza B viral hemagglutinin antigen that has an isolectric point greater than about 8.0 and is fused to at least one loop of domain 3 of the third flagellin.

[0013] In still another embodiment, the invention is a method of stimulating an immune response in a subject, comprising the step of administering to the subject a composition that includes at least four fusion proteins each of which activates a Tolllike Receptor 5, wherein: (a) a first fusion protein includes a first flagellin and a first influenza A viral hemagglutinin antigen that has an isoelectric point greater than about 6.0 fused to a portion of the first flagellin; (b) a second fusion protein includes a second flagellin and a second influenza A viral hemagglutinin antigen that is distinct from the first influenza A viral hemagglutinin antigen and that has an isoelectric point greater than about 7.0 fused to a portion of the second flagellin; (c) a third fusion protein includes a third flagellin and a first influenza B viral hemagglutinin antigen that has an isolectric point greater than about 8.0 and is fused to at least one loop of domain 3 of the third flagellin; and (d) a fourth fusion protein that activates a Toll-like Receptor 5 and includes a second influenza B viral hemagglutinin antigen that is distinct from the first influenza B viral hemagglutinin antigen and that has an isoelectric point greater than about 8.0 fused to at least one loop of domain 3 in the fourth flagellin.

[0014] The compositions and methods of the invention can be employed to stimulate an immune response, in particular, a protective immune response in the subject. Advantages of the claimed invention include, for example, the production of fusion proteins that have antigens with isoelectric points greater than about 7.0 that stimulate an adaptive immune response to the antigen sufficient to generate antibodies that provide protective immunity to a disease-causing organism that includes the antigen. The compositions of the invention provide an appropriate balance of innate and adaptive immune stimulating properties for use in vaccines that include antigens having an isoelectric point greater than about 7.0. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] FIG. 1 depicts the hemagglutinin (HA) trimer, the full length HA1 subunit monomer with an HA1 globular head subunit (membrane distal), and an HA2 membrane subunit (proximal stalk subunit). Portions of the HA1 subunit, referred to as "HAl-1" and "HA1-2" are depicted. The portion of HA referred to as "HA1-3" has been employed as a negative control and lacks sufficient secondary structure to essentially maintain the tertiary structure of the globular head of the portion of HA.

[0016] FIG. 2A shows that a fusion protein (SEQ ID NO: 122) that includes an R3 format of flagellin and a portion of HA (SEQ ID NO: 45), without a negatively charged linker (SEQ ID NO: 123) between the HA antigen and flagellin, did not induce an immunogenic response to the HA component of the fusion protein. Serum HAI antibody titers were measured by HAI test against B/Florida/4/06, and plotted as individual values with geometric mean titers (GMTs, horizontal lines) with 10 μg (·) or 3 μg (■) of HL199 (SEQ ID NO: 45) or with 10 μg (A) or 3 μg (▼) of HL098 (SEQ ID NO: 122). F147 formulation buffer was used a negative control (♦). BV (baculovirus) expressed HA0 (o, full-length HA) was used as a positive control.

[0017] FIG. 2B shows that the inclusion of an amino acid linker with a net negative charge (SEQ ID NO: 123) in the fusion protein depicted in FIG. 2 A (SEQ ID NO: 122) improved immunogenicity to the HA component of the fusion protein. Serum antibody titers were measured by MN test against B/Florida 4/06 virus, and plotted individually with GMTs (horizontal lines) with HL199 (·, SEQ ID NO: 145) or HL352 (T, SEQ ID NO: 125). Control groups included the formulation buffer (+, F147, negative) and BV (baculovirus) expressed HA0 (*, full-length HA, positive).

[0018] FIGs. 3A and 3B show in vivo TLR5 stimulated cytokine production: Comparison of fusion proteins for B/FL (Yamagata lineage) compositions: R3.HA1- 2 B (■, HL098, SEQ ID NO: 122) and R3.HA1-2 B with a 9 amino acid linker B (o, HL352, SEQ ID NO: 125). R3.HA1 (·, HL185, SEQ ID NO: 143) was included as a positive control. Dotted lines indicate (1) mean + 3 standard deviations (SD) of naive mice (....); (2) mean + 1 SD of HL077 ( , SEQ ID NO: 146); or (3) mean

+ 3 SD of HL185 (_._,·, SEQ ID NO: 143). IL-6 (FIG. 3A) and TNF (FIG. 3B) levels of individual mice are shown, along with bars representing the mean and standard error.

[0019] FIGs. 4A and 4B show in vivo TLR5 stimulated cytokine production: Comparison of fusion proteins for B/BR (Victoria lineage) compositions: R3.HA1-2 B (T, HL169, SEQ ID NO: 147) and R3.HA1-2 B with a 9 amino acid linker B (■, HL483, SEQ ID NO: 148). R3.HA1 (·, HL185, SEQ ID NO: 143) was included as a positive control. Dotted lines indicate (1) mean + 3 standard deviations (SD) of naive mice (--);(2) mean + 1 SD of HL077 ( , SEQ ID NO: 146); and (3) mean

+ 3 SD of HL185 (_._, ·, SEQ ID NO: 143). IL-6 (FIG. 4A) and TNF (FIG. 4B) levels of individual mice are shown, along with bars representing the mean and standard error.

[0020] FIGs. 5A and 5B show in vivo TLR5 stimulated cytokine production: Comparison of fusion protein compositions with additional negative charges.

Dotted lines indicate thresholds of cytokines levels correlated with immunogenicity. IL-6 (FIG. 5A) and TNF (FIG. 5B) levels of individual mice are shown, along with bars representing the mean and standard error. Formats for B/FL (Yamagata lineage) were compared: R3.HA1-2 B with a 9aa (amino acid) linker (SEQ ID NO: 123) B (■, HL352, SEQ ID NO: 125) and the same format with additional negative charges introduced in the linker (·, HL610, SEQ ID NO: 149). Formats for B/BR (Yamagata lineage) were compared: R3.HA1-2 B with a 9aa linker (SEQ ID NO: 123) B ( A . HL169, SEQ ID NO: 147) and the same format with additional negative charges introduced in the linker (♦, HL611, SEQ ID NO: 150). R3.HA1 (·, HL185, SEQ ID NO: 143) was included as a positive control and na^'fve (■) as a negative control.

[0021] FIG. 6 shows similar NIA activities of the two B/FL compositions that contain negatively charged amino acids in the linkers: B/Florida 4/06 R3 fusion protein compositions HL352 (o,SEQ ID NO: 125) and HL610 (D,SEQ ID NO: 149).

[0022] FIG. 7 shows a comparison of NIA activities of B/BR compositions with or without the use of additional negative charged amino acid residues (SEQ ID NO: 153) in the portion of HA that includes a portion of the globular head of HA:

B/Brisbane/60/08 R3 fusion protein compositions HL169 (o, SEQ ID NO: 147), HL483 (SEQ ID NO: 148) (□, with 9 amino acid linker of SEQ ID NO: 123), and HL61 1 (SEQ ID NO: 150) (Δ, with a 9 amino acid extension and other negatively charged amino acids of SEQ ID NO: 153).

[0023] FIGs. 8A and 8B show in vivo TLR5 stimulated cytokine production: Comparison of fusion protein compositions: B/FL (Yamagata lineage) R3.HA B (■, HL352, SEQ ID NO: 125) and D3Ins.HA B (A, HL656, SEQ ID NO: 151) and B/BR (Victoria lineage) R3.HA B (A , HL169, SEQ ID NO: 147) and D3Ins.HA B (T, HL657, SEQ ID NO: 128). R3.HA1 (·, HL185, SEQ ID NO: 143) was included as a positive control. Dotted lines indicate thresholds of cytokines levels correlated with immunogenicity. IL-6 (A) and TNF (B) levels of individual mice are shown, along with bars representing the mean and standard error.

[0024] FIGs. 9A through 9C show a comparison of NIA activities of fusion protein compositions of B/Florida/4/06 (o, HL352, SEQ ID NO: 125 and□, HL656, SEQ ID NO: 151) (A), B/Brisbane/60/08 (□, HL611, SEQ ID NO: 150 and o, HL657, SEQ ID NO: 128) (B), or B/Wisconsin/1/10 (A , HL724, SEQ ID NO: 152 and o, HL772, SEQ ID NO: 126) (C).

[0025] FIG. 10 shows D3Ins B/Fl composition elicits better MN titers than fusion proteins that include HA fused to R3 formats of flagellin: HL352 (SEQ ID NO: 125) (·, R3 with 9 an extension of SEQ ID NO: 123), HL610 (SEQ ID NO: 149) (o, R3 with 9 amino acid linker and additional negatively charged amino acids of SEQ ID NO: 153), or HL656 (SEQ ID NO: 151) (o, D3Ins with 9 amino acid linker of SEQ ID NO: 123). Data are shown as titers of individual mice with GMTs and seroconversion rates (percent (%) mice with a MN titer greater than about or equal to 80) above each group.

[0026] FIG. 1 1 shows relative immunogenicity of B/Wisconsin/1/2010 formats: R3 B/WI (·, HL719, SEQ ID NO: 154, R3 with negative linker), R3 B/WI (■, HL724, SEQ ID NO: 152, R3 construct format with additional negative charges in the amino acid extension), D3Ins B WI (A , HL772, SEQ ID NO: 126 or F147 buffer (·). Data are shown as titers of individual mice with GMTs. Statistical differences were determined in 1-way ANOVA/Tukey test. *,p < 0.05, ***,p < 0.001.

[0027] FIG. 12 shows HAI titers of monovalent flu B (influenza B)

compositions: R3 formats (HL611, SEQ ID NO: 150; HL753, SEQ ID NO: 155; HL742, SEQ ID NO: 156) or D3Ins format compositions of B/Brisbane/60/08-like viruses (B/Brisbane/60/08 or B/BR, HL657, SEQ ID NO: 128; B/Hong

Hong/259/10 or B/HK, HL774, SEQ ID NO: 157; and B/Bangladesh/5945/09 or B/BD, HL787, SEQ ID NO: 158). Controls included FLUVIRIN^® or F 147 buffer. Data are shown as titers of individual mice with GMTs. Statistical differences were determined in 2-way ANOVA/Tukey tests, *, p < 0.05, **,p < 0.01.

[0028] FIGs. 13 A and 13B show in vivo TLR5 stimulated cytokine production: Comparison of variations of D3Ins format fusion proteins: B Wisconsin (Yamagata lineage) D3I-ol (HL772, SEQ ID NO: 126), D3I-sl (HL848, SEQ ID NO: 160), and D3I-il (HL849, SEQ ID NO: 159). R3.HA1 (HL185, SEQ ID NO: 143) was included as a positive control.

[0029] FIG. 14 depicts the structure of flagellin, domains 0, 1, 2 and 3 and insertion sites in flagellin (SEQ ID NO: 2) for generation of fusion proteins that include antigens. Reference to amino acid numbers are made to SEQ ID NO: 2.

[0030] FIGs. 15A and 15B show the in vivo TLR5 stimulated cytokine production: Comparison of variations of D3I and D2I fusion protein constructs: B Wisconsin (Yamagata lineage) D3I-ol (HL772, SEQ ID NO: 126), D2I-ol (HL825, SEQ ID NO: 161), D2I-o2 (HL826, SEQ ID NO: 162), D2I-o3 (HL827, SEQ ID NO: 163), Dll-ol (HL828, SEQ ID NO: 164) and D2I-il (HL850, SEQ ID NO: 165) and B Brisbane (Victoria lineage) D3I-ol (HL657, SEQ ID NO: 128), D2I-ol (HL733, SEQ ID NO: 166), D2I-ol (HL856, SEQ ID NO: 167) and D2I-o2 (HL857, SEQ ID NO: 168). R3.HA1 (HL185, SEQ ID NO: 143) was included as a positive control. IL-6 (FIG. 15 A) and TNF-alpha (FIG. 15B) levels of individual mice are shown, along with bars representing the mean and standard error.

[0031] FIGs. 16A and 16B show in vivo TLR5 stimulated cytokine production: Comparison of variations of D3I and D2I fusion protein constructs: B Wisconsin (Yamagata lineage) were compared in FIG. 16A: D3I-ol (HL772, SEQ ID NO: 126), D3I-o2 (HL888, SEQ ID NO: 169), D2I-cl (HL890, SEQ ID NO: 170), D2I- i2 (HL892, SEQ ID NO: 171) and B Brisbane (Victoria lineage) were compared in FIG. 16B: D3I-ol (HL657, SEQ ID NO: 128), D2I-o3 (HL858, SEQ ID NO: 172), D3I-sl (HL860, SEQ ID NO: 173), D3I-o2 (HL889, SEQ ID NO: 174), D2I-cl (HL891, SEQ ID NO: 175) and D2I-i2 (HL893, SEQ ID NO: 176). R3.HA1 (HL185, SEQ ID NO: 143) was included as a positive control. The concentrations of IL6 from individual mice are shown, along with bars representing the mean and standard error.

[0032] FIG. 17 shows NIA assay for insertion variants (B/Wisconsin/1/2010): D3Ins (o, HL772, D3I-ol , SEQ ID NO: 126), various D2Ins (Δ, HL825, D2I-ol, SEQ ID NO: 161 ; 0, HL826, D2I-o2, SEQ ID NO: 162; ·, HL827, D2I-o3, SEQ ID NO: 163), and Dllns (■, HL828, Dll-ol, SEQ ID NO: 164), in comparison to R3 format (A , HL724, SEQ ID NO: 152).

[0033] FIGs. 18A and 18B show NIA assay for B/Wisconsin/1/2010 D2I variants in comparison to D3I format. A) B/Wisconsin/1/2010 D3I insertion fusion proteins: D3I-ol (o, HL772, SEQ ID NO: 126), D3I-sl (Δ, HL848, SEQ ID NO: 160), D3I-il (0, HL849, SEQ ID NO: 159), and D3I (A, HL864, SEQ ID NO: 177) in comparison to the D2I-il (·, HL850, SEQ ID NO: 165) and D2I (■, HL854, SEQ ID NO: 179) fusion proteins. B) B/Wisconsin/1/2010 D3I insertion fusion proteins: D3I-ol (o, HL772, SEQ ID NO: 126), D3I-o2 (□, HL888, SEQ ID NO: 169), and D3I-cl (Δ, HL890, SEQ ID NO: 170), in comparison to D2I-i2 (0, HL892, SEQ ID NO: 171).

[0034] FIGs. 19A shows NIA assay data for (B/Brisbane/60/2008) D3I fusion proteins: D3I-ol (o, HL657, SEQ ID NO: 128) and D3I-il (■, HL861, SEQ ID NO: 180), compared to the R3 (□, HL61 1, SEQ ID NO: 150) fusion protein.

[0035] FIG. 19B shows NIA assay for B/Brisbane/60/2008 R3 (□, HL611 , SEQ ID NO: 150) fusion protein in comparison to fusion protein insertion variants: D3I- ol (o, HL657, SEQ ID NO: 128), D2I-ol (Δ, HL733, SEQ ID NO: 166), D2I-ol (■, HL856, SEQ ID NO: 167), D2I-o2 (A , HL857, SEQ ID NO: 168), and D2I-il (♦, HL862, SEQ ID NO: 181). [0036] FIG. 20 shows NIA assay for B/Brisbane/60/2008 D3I fusion proteins: D3I-ol (o, HL657, SEQ ID NO: 128) and D2I-o3 (□, HL858, SEQ ID NO: 172), in comparison to D2I fusion proteins: D3I-sl (Δ, HL860, SEQ ID NO: 173), D3I-o2 (HL889, SEQ ID NO: 174), D2I-cl (HL891, SEQ ID NO: 175), D2I-i2 (HL893, SEQ ID NO: 176), and D2I-il (HL862, SEQ ID NO: 181).

[0037] FIGs. 21 A and 21B show HAI studies expressed as GMT in mice to treatment with A) B/Wisconsin/1/2010 (B/WI) fusion protein compositions comprising D3I-ol, HL772, SEQ ID NO: 126; D3I-il , HL849, SEQ ID NO: 159 and D2I-il, HL850, SEQ ID NO: 165 or B) B/Sichuan/379/1999 D3I-ol (HL863, SEQ ID NO: 182) or D3I-il(HL903, SEQ ID NO: 183). Positive (HA HU and HAO WI) and negative F147 (buffer) controls were also included. Seroconversion rates (percentage of mice show a HAI greater than about or equal to 40) are also given. ***, p < 0.001 vs F147 group in ANOVA test.

[0038] FIG. 22 shows HAI titers of mouse sera following treatment with B/Wisconsin/1/2010 (D3Ins B/WI, HL772, SEQ ID NO: 126); B/Hubei- Wujiagang/158/2009 (D3Ins B/HU HL869, SEQ ID NO: 184) and B/Texas/6/2011 (D3Ins B/TX HL871 , SEQ ID NO: 185). Horizontal lines represent GMTs

(numbers above). Seroconversion rates are shown.

[0039] FIG. 23 shows HAI titers elicited by D3Ins fusion protein compositions of B Brisbane or Bangladesh in rabbits. FLUVIRIN^® was included as a positive control, and F147 buffer was a negative control. Data are shown as titers of individual rabbits with bars. Horizontal lines represented GMTs (numbers above). Seroconversion percentage is shown.

[0040] FIG. 24 shows HAI titers of mouse sera following treatment with fusion protein composition STF2D3InsB/SI (HL863, SEQ ID NO: 182) or formulation buffer (F147). Data were plotted individually with GMTs and seroconversion indicated above.

[0041] FIGs. 25A and 25B show efficacy of treatment with fusion protein D3Ins B/SI 99 (HL863, SEQ ID NO: 182) compositions in mice challenged with

B/Sichuan/379/99 virus. Survival rates (FIG. 25A) and weights (mean percentage of initial weight) (FIG. 25B) are shown. [0042] FIG. 26 shows HAI Titers elicited by fusion proteins R3 B Wisconsin (HL724, SEQ ID NO: 152) and D3Ins B Wisconsin (HL772, SEQ ID NO: 126). CBER HA was included as a positive control and F147 buffer was included as a negative control. Data are shown as titers of individual rabbits with bars, and numbers representing geometric means and seroconversion percentages.

[0043] FIG. 27 shows HAI titers of rabbit sera following treatment with fusion protein STF2D3Ins B/WI (HL772, SEQ ID NO: 126), positive control Hubei virus, or formulation buffer (F147). Values are plotted individually with GMTs and seroconversion.

[0044] FIGs. 28 A-28C show reactogenicity of the HL772 fusion protein (SEQ ID NO: 126) or buffer control (F147) in rabbits. Groups of 10 rabbits were immunized once with the dose indicated on the x-axis. FIG. 28A depicts food consumption measured about 24 hours after immunization. FIG. 28B depicts temperature measured rectally about 6 hours post-immunization. FIG 28C depicts CRP was measured from serum taken about 24 hours after prime after the initial immunization (also referred to as "prime"). Data for all measures are shown as results of individual rabbits with lines representing means and standard error of the mean. Dotted lines represent the safety threshold calculated using the data from formula control rabbits.

[0045] FIG. 29 depicts loops in domain 3 for S. typhimurium FliC (SEQ ID NO: 1) based on a known crystal structure and predicted loops in S, typhimurium FljB (SEQ ID NO: 2).

[0046] FIG. 30 depicts predicted insertion sites in domains 0, 1, 2 and 3 of S. typhimurium FljB (SEQ ID NO: 2) for fusion with antigens, including insertion sites in a loop of domain 3. Domain 0 is predicted at amino acid residues 1-46 and amino acid residues 465-506; Domain 1 is predicted at amino acid residues 47-176 and amino acid residues 415-464; Domain 2 is predicted at amino acid residues 177-190 and amino acid residues 292-414; and Domain 3 is predicted at amino acid residues 191-291. The amino acid number of boundaries between domains 0, 1, 2 and 3 and insertion sites are also indicated. [0047] FIG. 31 depicts the domains (DO, D 1 , D2, D3) of a flagellin construct (SEQ ID NO: 283) and a fusion protein that includes, in sequence, the amino- domain 0, the amino-domain 1, the amino-domain 2, the carboxy-domain 2, the carboxy-domain 1 and the carboxy-domain 0 of flagellin. The antigen (Ag) is inserted between the amino-terminus and carboxy-terminus of domain 2 of the flagellin construct. The flagellin construct is referred to herein as "the 3 construct" (SEQ ID NO: 283) and lacks a domain 3, which is present in the naturally occurring flagellin (SEQ ID NO: 2).

[0048] FIG. 32 depicts the domains (DO, Dl, D2, D3) of a flagellin construct (SEQ ID NO: 284) and a fusion protein that includes, in sequence, the amino- domain 0, the amino-domain 1, the amino-domain 2, the carboxy-domain 2, the carboxy-domain 1 and the carboxy-domain 0 of the flagellin. Two antigens (Ag) are in the fusion protein. An antigen is fused between the amino-domain 2 and the carboxy-domain 2 of the flagellin construct. Another antigen is fused to the carboxy-terminal amino acid of the Domain 0 of the flagellin construct. The flagellin construct is referred to herein as "the R3-2xAg" construct, which lacks a domain 3 of the naturally occurring flagellin (SEQ ID NO: 2).

DETAILED DESCRIPTION OF THE INVENTION

[0049] The features and other details of the invention, either as steps of the invention or as combinations of parts of the invention, will now be more particularly described and pointed out in the claims. It will be understood that the particular embodiments of the invention are shown by way of illustration and not as limitations of the invention. The principle features of this invention can be employed in various embodiments without departing from the scope of the invention.

[0050] The invention is generally directed to compositions that include fusion proteins having an antigen with an isoelectric point greater than about 7.0 that activate Toll-like Receptor 5 and use of the fusion proteins in compositions to stimulate immune responses in a subject, such as protective immune responses.

[0051] A threshold of TLR5 signaling must be reached to promote the immunogenicity of fusion proteins that include flagellin and antigens. However, excess TLR5 signaling heightens proinflammatory cytokine response that can elicit systemic reactogenicity (e.g., unwanted side-effects, such as fever, weight loss) in vivo. Fusion proteins that include flagellin must activate TLR5 signaling to promote an adaptive immune response to the antigen with an isoelectric point greater than about 7.0, while simultaneously remaining below a threshold of TLR5 signaling that results in unacceptable reactogenicity.

[0052] Mouse and rabbit animal models to assess the minimal TLR5 activity required for a fusion protein that includes flagellin to be immunogenic (mouse model) and the maximal TLR5 activity associated with a well tolerated composition that includes a fusion protein of flagellin for use in a vaccine (rabbit model) have been developed, as described herein. In the mouse TLR5 activity model, when a mouse is immunized with about 1 μg of a fusion protein, the fusion protein must elicit at least about 32 pg/ml of IL6 and about 80 pg/ml of TNF in the serum about 3 hours post immunization to be considered immunogenic. In the rabbit model, a dose that elicits levels of food consumption that are greater than about 1 standard deviation below the mean of a buffer control, rises in body temperature that are greater than about 3 standard deviations away from the mean of the control and rises in C-reactive protein (CRP) that are greater than about 10 standard deviations away from the mean of the buffer control are likely to be reactogenic (e.g., have negative, unwanted side effects) in a human. The dose range for which a composition of the invention would be considered useful as a vaccine remains non-reactogenic in the rabbit model and is described as the "safety window." The dose range for which a composition of the invention would be considered useful as a vaccine is

immunogenic, yet remains non-reactogenic in preclinical and clinical models is described as the "therapeutic window." The fusion proteins properly balance activation of TLR5 signaling and adaptive immune responses to the antigen to achieve a therapeutic window that is acceptable for use in methods to stimulate protective immunity.

[0053] Full length flagellin, without fusion to any antigen, or flagellin fused to an antigen at the carboxy-terminus of the flagellin have the narrowest safety window for use in compositions to stimulate immune responses in subjects because such compositions trigger a maximal TLR5 signaling response. Antigens with an isoelectric point less than about 7.0 fused to R3 constructs of flagellin, in which domain 3 of flagellin is entirely replaced with an antigen (U.S. Application No: 12/905,584), have a wider safety window, believed to be due, in part, to the ability of the antigen to sterically hinder the fusion protein binding to TLR5 to thereby rendering it a "partial agonist." Partial agonist activity of R3 flagellin fusion proteins have been described for A/Puerto Rico/08/34 and A/California/07/09 (Taylor, et al, Vaccine 30:5761-5769 (2012)).

[0054] There must be sufficient partial TLR5 agonist activity for a fusion protein that includes flagellin to generate an immune response to the antigen component. As described herein, antigens with isoelectric points greater than about 7.0 when fused to R3 and R32x constructs of flagellin, result in decreased TLR5 signaling in the mouse model of TLR5 activity (decreased stimulation of adaptive immune responses), and, thus, are poorly immunogenic. Antigens with isolectric points greater than about 7.0 may result in intramolecular interactions between the negatively charged flagellin and the antigen. Fusion of an antigen with an isoelectric point greater than about 7.0 (e.g., about 7.0, about 7.5, about 8.0, about 8.5, about 9.0, about 9.5, about 10.0, about 10.5, about 11.0, about 11.5) to at least one loop in domain 3 of flagellin, is believed to form a configuration that maximizes the distance of the antigen from the TLR5 binding domain of flagellin to thereby preserve the ability of the fusion protein to activate TLR5 by flagellin binding to TLR5.

[0055] In an embodiment, the invention includes a fusion protein comprising a flagellin and at least one antigen that has an isoelectric point greater than about 7.0 and that is fused to at least one loop of domain 3 of the flagellin. The fusion protein of the invention activates the Toll-like Receptor 5. In an embodiment, an antigen of the fusion proteins of the invention and employed in the methods of the invention have an isoelectric point greater than about 7.5.

[0056] In another embodiment, an antigen of the fusion proteins of the invention and fusion proteins employed in the methods of the invention has an isoelectric greater than aobut 8.0. [0057] In still another embodiment, an antigen of the fusion proteins of the invention and fusion proteins employed in the methods of the invention has an isoelectric greater than about 8.5.

[0058] In a further embodiment, an antigen of the fusion proteins of the invention and fusion proteins employed in the methods of the invention has an isoelectric greater than about 9.0.

[0059] In still another embodiment, an antigen of the fusion proteins of the invention and fusion proteins employed in the methods of the invention has an isoelectric greater than about 9.5.

[0060] In an additional embodiment, an antigen of the fusion proteins of the invention and fusion proteins employed in the methods of the invention has an isoelectric greater than about 10.0.

[0061] In another embodiment, an antigen of the fusion proteins of the invention and fusion proteins employed in the methods of the invention has an isoelectric greater than about 10.5.

[0062] In yet another embodiment, an antigen of the fusion proteins of the invention and fusion proteins employed in the methods of the invention has an isoelectric greater than about 11.0.

[0063] Fusion of the antigen to a loop of domain 3 of flagellin essentially retains domain 3 of flagellin in its tertiary structure. The phrase "essentially retains domain 3 of flagellin in its tertiary structure," as used herein, refers to maintenance of the tertiary structure of domain 3 of flagellin, which can be assessed by well-established in vivo and in vitro assays described herein that are known to one of ordinary skill in the art, including the ability of flagellin to activate TLR5 and to assess protective immunity.

[0064] "Fusion protein," as used herein, refers to a protein generated from at least two distinct components, a flagellin or a portion of a flagellin and an antigen or a portion of an antigen having an isoelectric point greater than or equal to 7.0, including antigens with an isoelectric point of at least one member selected from the group consisting of about 7.0, about 7.5, about 8.0, about 8.5, about 9.0, about 9.5, about 10, about 10.5 and about 11.0. [0065] Fusion proteins of the invention can be generated recombinantly or by chemical conjugation using well-established techniques. A recombinant fusion protein can be generated by operably linking a nucleic acid sequence encoding a flagellin, or a portion of a flagellin that includes a domain 3, to a nucleic acid sequence encoding an antigen, such as an antigen that is at least a portion of an influenza viral antigen. Fusion proteins of the invention can include, for example, one, two, three, four or five, antigens fused to, for example, one, two, three, or four loops of domain 3 of one or more flagellin.

[0066] The isoelectric point (pi) of a protein is the pH at which a particular antigen carries no net charge. The isoelectric point of the antigen fused to the flagellin can be at least one member selected from the group consisting of about 7.0, about 7.5, about 8.0, about 8.5, about 9.0, about 9.5, about 10.0, about 10.5 and about 11.0. In an embodiment, the pi of the antigen is greater than about 7.0. In another embodiment, the pi of the antigen is greater than about 7.5. In a further embodiment, the pi of the antigen is greater than about 8.0. In still another embodiment, the pi of the antigen is greater than about 8.5. In yet another embodiment, the pi of the antigen is greater than about 9.0. In an additional embodiment, the pi of the antigen is greater than about 9.5. In another embodiment, the pi of the antigen is greater than about 10.0. In yet another embodiment, the pi of the antigen is greater than about 10.5. In still another embodiment, the pi of the antigen is greater than about 11.0.

[0067] The flagellin in the fusion proteins of the invention can be an S.

typhimurium flagellin (UniProt accession number P06179 or P52616), such as an S. typhimurium flagellin selected from the group consisting of SEQ ID NOS: 1 and 2; . E, coli flagellin (UniProt accession number A0PCV8), such as, for example, SEQ ID NO: 3; P. aeruginosa flagellin (UniProt accession number P72151), such as SEQ ID NO: 4; Aquifex aeolicus VF5 flagellin (UniProt accession number 067803), such as SEQ ID NO: 5; Helicobacter pylori J99 Flagellin A (UniProt accession number P0A052), such as SEQ ID NO: 6; and Legionella pneumophila flagellin (UniProt accession number Q48824), such as SEQ ID NO: 7. [0068] The flagellin employed in the fusion proteins of the invention can lack at least a portion of a carboxy-domain 0 or a portion of an amino-domain 0, for example, at least one member selected from the group consisting of about 5, about 10, about 14, about 15 and about 20 amino acids of the carboxy-domain 0 or a portion of an amino-domain 0. As shown in FIG. 14, flagellin includes a carboxy- domain 0, a carboxy-domain 1, a carboxy-domain 2, a domain 3, an amino-domain 2, an amino-domain 1 , and an amino-domain 0. When flagellin assumes its tertiary structure, domains 2 and 3 form a juncture in which the flagellin folds on itself so that the most carboxy-amino acid of the carboxy-domain 0 of flagellin is adjacent to the most amino-amino acid of the amino-domain 0 of flagellin. Crystallographic studies of flagellin have identified the secondary and tertiary structure of S.

typhimurium FliC flagellin (SEQ ID NO: 1) (PDB code: lUCU).

[0069] The phrase "a loop of domain 3 of flagellin," as used herein, refers to a stretch of amino acids within domain 3 of flagellin that is, itself, devoid of secondary structures (e.g., B-sheets, a-helices), yet flanks adjacent stretches of amino acids in domain 3 that include secondary structures, such as B-sheets, a- helices. Loops of domain 3 in flagellin can be about 2, about 3, about 4, about 5, about 6, about 7 and between about 5 to about 30 amino acids in length. Exemplary loops of domain 3 of S. typhimurium flagellin (SEQ ID NO: 1) are depicted in FIGS. 14, 29 and 30 and are referred to herein as "D3I-ol," "D3I-il," "D3I-cl" and D3I-sl."

[0070] Flagellin (FljB) from Salmonella typhimurium is depicted in SEQ ID NO: 2. Domain 3 of Salmonella typhimurium flagellin is between amino acid residue 191 and amino acid residue 291 of SEQ ID NO: 2. Flagellin from E. coli (UniProt accession number A0PCV8) is depicted in SEQ ID NO: 3. Domain 3 of E. coli flagellin of SEQ ID NO: 3 is predicted between amino acid residue 191 and amino acid residue 283 of SEQ ID NO. 3. P. aeruginosa flagellin (UniProt accession number P72151) is depicted in SEQ ID NO: 4 with domain 3 predicted between amino acid residue . Flagellin from Aquifex aeolicus VF5 (UniProt accession number 067803) is depicted in SEQ ID NO: 5. Domain 3 of Aquifex aeolicus flagellin is predicted between amino acid residue 197 and amino acid residue 302 of SEQ ID NO. 5. The flagellin A from Helicobacter pylori J99 (UniProt accession number P0A052) is depicted in SEQ ID NO: 6. Domain 3 of Helicobacter pylori J99 of SEQ ID NO: 6 is predicted between amino acid residue 189 and amino acid residue 283 of SEQ ID NO: 6. The flagellin from Legionella pneumophila (UniProt accession number Q48824) is depicted in SEQ ID NO: 7. Domain 3 of Legionella pneumophila flagellin of SEQ ID NO: 7 is predicted between amino acid residue 189 and amino acid residue 283 of SEQ ID NO: 7.

[0071] X-ray crystallography of Salmonella typhimurium FliC flagellin (SEQ ID NO: 1) shows that domain 3 of flagellin includes 6 loops (FIG. 29). The loops in domain 3 of Salmonella typhimurium flagellin SEQ ID NO: 1 are from amino acid residues 211 to 212 (loop 1); amino acid residues 217 to 219 (loop 2); amino acid residues 223 to 229 (loop 3); amino acid residues 237 to 242 (loop 4); amino acid residues 250 to 255 (loop 5) and amino acid residues 259 to 275 (loop 6).

[0072] FIG. 29 identifies the predicted loops (gray) in domain 3 of Salmonella typhimurium FljB flagellin (SEQ ID NO: 2) and compares the location in the sequence to the loops in domain 3 of Salmonella typhimurium FliC flagellin (SEQ ID NO: 1). Darkly shaded arrows depict secondary structures in domain 3 of S. typhimurium FliC, such as β-sheets and a-helices.

[0073] FIG. 30 depicts the amino acid sequence of Salmonella typhimurium FljB flagellin (SEQ ID NO: 2), boundaries of domain 0, 1, 2 and 3, and potential sites of fusion of antigens having isoelectric points greater than about 7.0 in loops of domain 3 of the flagellin. In an embodiment, the site of fusion of an antigen having an isoelectric point greater than about 7.0 is D3I-il (also referred to as "D3Ins-il") between amino acid residues 259 and 260 of SEQ ID NO: 2 in loop 5 of domain 3 (see FIG. 29). In another embodiment, the site of fusion of an antigen having an isoelectric point greater than about 7.0 is D3I-ol (also referred to as "D3Ins-ol") between amino acid residues 277 and 278 of SEQ ID NO: 2 in loop 6 of domain 3 (see FIG. 29).

[0074] In another embodiment, the D3I-ol site of fusion of an antigen having an isoelectric point greater than about 7.0 is between amino acid residues 274 and 275 of SEQ ID NO: 4 in predicted loop 6 of domain 3. In still another embodiment, the D3I-ol site of fusion of an antigen having an isoelectric point greater than about 7.0 is between amino acid residues 274 and 275 of SEQ ID NO: 5 in predicted loop 6 of domain 3.

[0075] In another embodiment, the D3I-il site of fusion of an antigen having an isoelectric point greater than about 7.0 is between amino acid residues 258 and 259 of SEQ ID NO: 4 in predicted loop 5 of domain 3. In still another embodiment, the D3I-ol site of fusion of an antigen having an isoelectric point greater than about 7.0 is between amino acid residues 260 and 261 of SEQ ID NO: 5 in predicted loop 5 of domain 3.

[0076] The antigen can be fused to domain 3 of flagellin to generate constructs referred to as D3I-ol and D3I-il . "D3I-ol," as used herein, refers to insertion into a loop of domain 3 (Domain 3 Insertion) on the outer (o) or concave surface of flagellin, such as loop 6 OF SEQ ID NO: 2 or SEQ ID NO: 1. "D3I-il," as used herein, refers to a loop of domain 3 on the inner (i) or convex surface of flagellin, such as loop 3 of SEQ ID NO: 2 and SEQ ID NO: 1. The tertiary structure of flagellin and portions of flagellin that would be considered concave and convex surfaces of flagellin are described, for example, by Samatey, et at, Nature 410:331- 337 (2001).

[0077] The designation "c," with respect to an insertion of an antigen into a loop of domain 3 of flagellin (D3I-cl), refers to a "side-way" portion of domain 3 (FIG. 14).

[0078] The designation "s," with respect to an insertion of an antigen into a loop of domain 3 of flagellin (D3I-sl), means the "tip" of domain 3 of flagellin (FIG. 14).

[0079] Exemplary fusion proteins of the invention are shown in SEQ ID NOS: 126, 128-130, 151, 157-193.

[0080] The junction of the flagellin component and antigen component of fusion proteins of the invention results in a sequence of unique amino acids. For fusion proteins employed in methods of the invention to treat humans, if this unique sequence of amino acids at the juncture of the fusion of the flagellin component and the antigen component shares homology with a known human protein, the fusion protein has the potential to elicit an unwanted immune response to the human protein or a portion of the human protein. Upon selection of an insertion site in flagellin (e.g., a loop of domain 3, an adjacent portion of the carboxy- or amino- terminus of domain 2 of an R3 construct), the sequence of unique amino acids that would be created by fusion of the antigen to flagellin is assessed for its potential ability to elicit an unwanted immune response. In the evaluation, a probe of about 10 to about 12 amino acids in length, which includes the flagellin/antigen junction is used to probe a database of known human genome sequences. If homology is identified for a stretch of amino acids greater than about 5 amino acids, then the junction sequence is modified with an amino acid substitution of, for example, 1, 2, 3, 4, 5, or 6 amino acids to decrease the homology. The order of preferred amino acids for use in the modification is serine, threonine, alanine and glycine. Generally, a single amino acid substitution is sufficient to modify the homology. For example see SEQ ID NO: 283 for the native FljB sequence without domain 3. For example, in an R3 construct of flagellin, Arg405 of SEQ ID NO: 283 can be modified to an alanine residue to generate SEQ ID NO: 284. Alternatively, Alal91 of SEQ ID NO: 283 can be modified to a serine residue, Arg405 of SEQ ID NO: 283 can be modified to an alanine and a serine residue can be added to the modified alanine residue (i.e., the most carboxy-terminus amino acid) to generate SEQ ID NO: 285. An exemplary fusion protein with modified amino acid residues at the junction of fusion of the antigen and the flagellin is SEQ ID NO: 268, where an HA1-2 portion of SEQ ID NO: 286 is fused to a flagellin of SEQ ID NO: 285. In SEQ ID NO: 268, the junctions of the fusion of the antigen and the flagellin are at Alal90 and Glyl91; Ser413 and Ser414; and Ala628 and Ser629 with Gly630.

[0081] In an embodiment, the flagellin employed in the fusion protein of the invention lacks at least one member selected from the group consisting of at least a portion of the carboxy-domain 0 and at least a portion of an amino-domain 0.

[0082] In another embodiment at least one additional antigen is fused to the flagellin at a site that is distinct from fusion of the antigen to a loop of domain 3 of the flagellin. The additional antigen can be an antigen similar to the antigen fused to at least one loop of domain 3 of the flagellin. Alternatively, the additional antigen can be an antigen that is different (also referred to herein as "distinct") from the antigen that is fused to at least one loop of domain 3 of the flagellin. For example, at least a portion of an influenza hemagglutinin antigen, such as an HAl-1 or an HA1-2 portion, can be fused to a loop of domain 3 (e.g., between amino acid 277 and amino acid 278 of loop 3 of SEQ ID NO: 2) of flagellin and a different portion of an influenza hemagglutinin antigen, such as HA1-1L, can be fused to a different loop of domain 3, such as between amino acids 259 and 260 of loop 5 of SEQ ID NO: 2.

[0083] In another embodiment, the site of insertion of an antigen in a loop of domain 3 of flagellin can be 1, about 2, about 3, about 4, about 5, about 6, about 7, about 8, about 9, about 10, amino acids towards the carboxy-domain 2 or an amino- domain 2 of flagellin from the insertion sites identified in FIGs. 14, 29 and 30. For example, with reference to SEQ ID NO: 2, a loop of domain 3 to which the antigen is fused is between amino acid residues 277 and 278 (e.g., D3I-ol) (FIG. 30).

Alternatively, fusion of the antigen can occur between amino acid 266 and 282 of SEQ ID NO: 2.

[0084] Likewise, with reference to SEQ ID NO: 2, a loop of domain 3 to which the antigen is fused is between amino acid residues 259 to 260 (D3I-il) or amino acid residues 260 to 261. The insertion site in a loop of domain 3 in the D3I-il fusion protein of SEQ ID NO: 2 is between amino acid residues 190 to 191 or between amino acid residues 291 to 292.

[0085] In another embodiment, about 2 to about 4 amino acid residues in the loop of domain 3 can be deleted prior to fusion with the antigen with an isoelectric point greater than about 7.0. The deletions would be designed so that adjacent secondary structures in the flagellin would not be disrupted. Such deletions in at least one loop of domain 3 of flagellin may be employed when fusing flagellin to relatively large antigens.

[0086] In a preferred embodiment, the flagellin for use in the fusion proteins is a flagellin that includes at least one member selected from the group consisting of SEQ ID NOS: 1-7. The antigen is fused between amino acid residue 191 and amino acid residue 285 of SEQ ID NO: 1, which is within a loop of domain 3 of flagellin.

[0087] In a particular embodiment, the antigen that is fused to a loop of domain 3 of flagellin is an influenza viral antigen, in particular, an influenza B viral antigen. [0088] In another embodiment the influenza viral antigen is at least a portion of an influenza A antigen subtype (H3, H7, H5 or H9). The influenza A antigen subtype can be at least one member selected from the group consisting of the H3, H5, H7 and H9. An additional influenza A antigen subtype can be at least one member selected from the group consisting of an HI and an H2 subtype, (HI subtype or H2 subtype).

[0089] In a particular embodiment, the influenza viral antigen is hemagglutinin, in particular, a portion of a hemagglutinin antigen that includes at least a portion of a globular head of the hemagglutinin antigen. The portion of an influenza viral hemagglutinin employed in the fusion proteins of the invention preferably includes at least a portion of a globular head of hemagglutinin that includes a sialic acid binding site. The portion of the globular head of the hemagglutinin includes at least one β-sheet at the bottom of the globular head that causes the globular head to essentially retain its tertiary structure.

[0090] "A sialic acid binding site," as that phrase is used herein in reference to the portion of the protein from the naturally occurring viral hemagglutinin, means a part of the influenza viral hemagglutinin that has the capacity to interact with sialic acid residues. "A sialic acid binding site" is also referred to herein as "a sialic acid binding domain."

[0091] "At least a portion," as used herein, refers to any part of the antigen or flagellin that is less than the entirety of the antigen or flagellin.

[0092] "A globular head," as that phrase is used herein, refers to a portion of a protein of an influenza viral hemagglutinin that includes the receptor or sialic acid binding regions. "Globular head," is also referred to herein as a "globular domain." The globular head of viral hemagglutinin proteins has been determined based on x- ray crystallography as described, for example, by Wilson I.A., et al Nature

259:366-373 (1981); Chen, J., et al, Cell 95:409-417 (1998); Ha Y., et al , The EMBO Journal 27:865-875 (2002); Russell, R.J., et al , Virology 525:287-296 (2004); and Cox, N.J., et al, In: Toply and Wilson's Microbiology and Microbial Infections, eds. BWJ Mathy, et al., Vol. 1 (9^th ed.) New York, NY, Oxford Univ. Press, Ch. 32, p. 634 (1998). The globular head of an influenza viral hemagglutinin is a component of the HA1 subunit of influenza viral hemagglutinin. In addition to the receptor binding domain, the globular head can include the E^'subdomain and F^" subdomain as described, for example, by Ha, Y., et al. The EMBO Journal 27:865- 875 (2002).

[0093] The phrase, "causes the globular head to essentially retain its tertiary structure," as used herein, refers to maintenance of the tertiary structure of the globular head to thereby mimic the tertiary structure of the globular head in the naturally occurring influenza viral hemagglutinin, which can be assessed by the ability to generate a sufficient immune response to stimulate a protective immune response in a subject in vivo or viral neutralization in in vitro assays, as described herein.

[0094] Influenza viruses are single-stranded RNA viruses that belong to the viral family Orthomyxoviridae. Influenza viruses are divided into three types (A, B, C) determined by the antigenic differences in ribonucleoprotein (RNP) and matrix (M) antigens of the viruses. Influenza A virus naturally infects humans and several other mammalian species, including swine and horses, and a wide variety of avian species, and causes epidemics and pandemics in the human population. Influenza B virus appears to naturally infect only humans and seals and can cause epidemics in humans. Influenza C virus has been isolated from humans and swine, but generally does not occur in epidemics and usually results in mild disease in humans.

[0095] Mature influenza virions are enveloped with a pleomorphic structure ranging in diameter from about 80 to about 120 nm. The single-stranded RNA genome is closely associated with a helical nucleoprotein and is present in seven (influenza C) or eight (influenza A and B) separate segments of ribonucleoprotein (RNP), each of which has to be present for successful replication of the virus. The segmented genome is enclosed within an outer lipoprotein envelope. Matrix protein 1 (MP1 or also referred to herein as "Ml") lines the inside of the outer lipoprotein envelope and is bound to the RNP.

[0096] Hemagglutinin (HA) is a surface glycoprotein on a virus (e.g., an influenza virus) that is responsible for binding to N-AcetylNeuraminic Acid

(NeuNAc; also referred to herein as "sialic acid") on host cells and subsequent fusion of viral and host membranes. HA acquired its name by virtue of its ability to cause red blood cells to clump, or agglutinate. Influenza HA consists of the three monomeric (HA0) subunits. HA performs two critical functions during the infection process: binding to a cell surface sialyloligosaccharide receptor and fusion of virus and host cell membrane. Following binding of the HA to the plasma membrane of a host cell, the host cell membrane engulfs the virus in an endosome and attempts to digest the contents of the endosome by acidifying its interior and transferring it to a lysosome in the host cell. However, the acidic environment of the lysosome destabilizes HA, resulting in partial unfolding of HA0 which exposes a protease- sensitive site (the maturational cleavage site) that is cleaved by a host protease to form HAl and HA2 subunits which are connected by a single disulfide bond (Wiley, D. C, et al. , Annu. Rev. Biochem. 5(5:365-394 (1987)). Cleavage occurs at a specific amino acid residue and generates a hydrophobic amino terminus for the HA2 subunit. This hydrophobic terminus of HA2 mediates fusion between the viral envelope and the endosomal membrane of the host cell and releases the contents of the virion into the cytoplasm of the cell, a process known as uncoating. Thus, cleavage of the HA polypeptide is a requirement for infectivity.

[0097] The crystal structure of several viral hemagglutinins has been determined (see, for example, Wilson, I.A., et al, Nature 259:366-373 (1981); Chen, J., et al, Cell 95:409-417 (1998); Ha, Y., et al, The EMBO Journal 21: 865-875 (2002); Russell, R.J., et al, Virology 325:287-296 (2004); and Cox, N.J., et al, In: Toply and Wilson's Microbiology and Microbial Infections, eds. B.W.J. Mathy, et al, Vol. 1 (9^th ed.) New York, NY, Oxford Univ. Press, Ch. 32, p. 634 (1998)). X-ray crystallographic structures show that HA is folded into two structural components or domains - a globular head and a fibrous stalk (see, for example, Figure 1). The globular head includes HAl , including that part of HAl that binds to sialic acid (also referred to as the "receptor binding site or domain" or "sialic acid binding site or domain"), and antiparallel β-sheets. The fibrous stalk is more proximal to the viral membrane and includes the HA2 subunit and part of HAl, including the cleavage site between HAl and HA2. [0098] There are seventeen known subtypes of Influenza A HA (HI -HI 7) that share between about 40 to about 60% sequence identity (Tong, et ah, Proc. Natl. Acad. Sci. , 109:4269-42 '4 (2012)). Influenza viruses containing all 17 HA subtypes have been isolated from, for example, avian species (H5, H7, and H9), equine (H3 and H7), seals (H3, H4 and H7), whales (HI and H13) and swine (HI, H3, and H9). Subtypes of influenza A virus are generally named according to the particular antigenic determinants of HA (H, 17 major types) and neuraminidase (N, about 9 major types). For example, subtypes include influenza A (H2N1), A(H3N2), A(H5N1), A(H7N2), A(H9N2), A(H1/H0), A(H3/H0), A(H5/H0) and A(H7N9). In the last century, three subtypes of influenza A resulted in pandemics: HI in 1918, 1977 and 2007; H2 in 1957 and H3 in 1968. In 1997, an H5 avian virus and in 1999, an H9 virus resulted in outbreaks of respiratory disease in Hong Kong.

[0099] HA from influenza type B viruses have been isolated from humans and seals and are not divided into subtypes, although influenza type B viruses are characterized by two antigenically different lineages, Yamagata and Victoria.

Influenza type B virus strains from both the Yamagat and Victoria lineages typically co-circulate. Compositions used in influenza vaccines can include four antigens (i.e., "quadrivalent flu vaccines") that, in addition to the two A strains, include both the circulating B strains. Similar to influenza A HAs, X ray crystallagraphic studies of influenza B HAs have revealed four major antigenic regions that are located in the vicinity of the receptor binding site, which is in the globular head of the influenza B HA (Wang, Q., et al, J. Virol , March:301 1-3020 (2008)).

[00100] A host infected with influenza can mount an antibody response to the globular head of HA that protects that host from subsequent infection with the same strain of virus by blocking the interaction between HA and the host cell, i.e., neutralizing the infectivity of the virus. Due to the low fidelity and high rate of influenza RNA replication, the virus is constantly experiencing minor mutations in the HA gene that preserve the globular head structure and host cell interaction, but may allow progeny virus to escape immune surveillance. These point mutations are referred to as "antigenic drift." In addition, if a single host is simultaneously infected with two different strains of influenza A, a new subtype of virus may emerge as a result of reassortment, or the exchange of the RNA segments, or genes, between different strains of influenza A viruses. The viruses emerging from reassortment present the human immune system with a new antigenic experience that usually results in high morbidity and mortality. This type of drastic antigenic change is known as "antigenic shift." Since type B influenza viruses circulate almost exclusively in humans, these viruses cannot undergo reassortment with animal strains and, thus, are changed only by antigenic drift.

[00101] Immunity to HA can reduce the likelihood of infection and severity of disease if infection does occur. HA is an important antigenic target and the efficacy of vaccines depends on the antigenic match between the vaccine strain and the circulating strain. Since the hemagglutinin protein readily undergoes antigenic shift and drift in order to evade the host's immune defense, traditional vaccines must be based on currently circulating influenza strains and annually updated. Annual updates of influenza vaccines are not only costly they also require significant amounts of production time and manufacturing infrastructure. A vaccine composition based on invariant regions of the virus may provide broadly cross- reactive protection.

[00102] Hemagglutinin forms a trimer on the surface of the influenza virus and infected cell. The monomeric subunits of the trimer are depicted in FIG. 1, and include an HA1 subunit (membrane distal globular head) and HA2 subunit

(membrane proximal "stalk"). Portions of HA for use in the compositions of the inventor can include at least a portion of the globular head, which includes the cell surface receptor binding site and the majority of the neutralizing antibody epitopes. Portions of HA, referred to as HAl-1 and HA 1-2 subunits, fused to flagellin (R3 and Pv32Xconstructs of flagellin, see, for example, U.S. Patent Application No.:

12/905,584), have previously been employed in compositions that provide protective immunity to viral challenge (see, for example, U.S. Patent No.: 8,420,102). The HAl-1 and HA 1-2 portions of HA include a portion of the globular head and differ in amino acid length and secondary structure. The HAl-1 and HA 1-2 portions include at least a portion of the globular head of HA and at least one β-strand at the bottom of the portion of the globular head that maintains the portion of the globular head in its tertiary structure.

[00103] When fused to R3 or R32x formats (also referred to herein as "R3 constructs" or "R32x constructs") of flagellin, HAl-1 and HA1-2 portions of certain influenza viral antigens can provide protective immunity against viral challenge (see, for example, U.S. Patent No.: 8,420,102). Fusion proteins that include flagellin lacking the entirety of domain 3 and in which an HA antigen has been fused to the flagellin in the region of flagellin that was domain 3 are referred to as "R3 fusion proteins." "R32x fusion proteins (also referred to "R3.2x")" are R3 fusion proteins that include a second antigen, which is the same or distinct from the antigen in the region of flagellin that was domain 3, fused to the most carboxy -terminal amino acid of the carboxy-domain of flagellin (see, for example, U.S. Patent Application No. 12/905,584). For example, an HAl-1 portion can be fused in the region of flagellin that was domain 3 and an HA 1-2 portion can be fused to the most carboxy -terminal amino acid of flagellin to form an R32x fusion protein.

[00104] In contrast, in certain influenza A strains, R3 and R3.2x fusion proteins that include portions of an influenza B HA, are poorly immunogenic, as shown herein. For example, influenza B fusion proteins that include an HA1-2 portion (SEQ ID NO: 45) of the Yamagata lineage strain B/Florida/4/2006, fused to flagellin in an R3 format (HL098, SEQ ID NO: 122) or an R2X format (HL1 18, SEQ ID NO: 287) or Victoria lineage strain B/Brisbane/60/2008, SEQ ID NO: 264 fused to flagellin in an R3format (HL169, SEQ ID NO: 147) or an R32X format (HL171, SEQ ID NO: 288) failed to elicit measurable levels of hemagglutination inhibition (HAI) titers, a standard measure of protective immunity against influenza virus. Every year, new multivalent blends of compositions for use as vaccines are developed as treatments to prevent and manage disease associated with viral influenza infection. For example, trivalent influenza vaccine (TIV) treatments contain two different inactivated influenza type A strains and one inactive influenza type B strain, e.g. FLUVIRIN^®. Quadrivalent vaccine (QIV) treatments are available, e.g. FLUZONE^® Quadrivalent vaccine, which contains two different inactivated influenza type A strains and two inactivated influenza B strains. Therefore, a need exists to develop new and effective, fusion proteins targeting influenza B that can be included in multivalent compositions.

[00105] Influenza viral antigens fused to at least one loop of domain 3 of flagellin can be a portion of an influenza B viral antigen, such as a portion referred to herein as "HAl-l," "HA1-2" and "HA1-1L." Compositions that include influenza viral hemagglutinn antigens, such as HAl-1, HA 1-2 and HA1-1L portions of influenza hemagglutinin, can be employed in methods to treat or prevent seasonal and pandemic influenza in subjects.

[00106] "A portion," as used herein with reference to an influenza viral antigen, for example, an influenza viral hemagglutinin antigen, refers to any part of the influenza viral hemagglutinin that is less than the entirety of the influenza viral hemagglutinin.

[00107] In an embodiment, the portion iof the influenza viral hemagglutinin (HA) antigen is a portion of the globular head of HA. In another embodiment, the portion of the influenza viral HA is the entirety of the globular head.

[00108] "HAl -1 ," as used herein, refers to a protein portion of a viral hemagglutinin that includes at least about one β-sandwich that includes the substrate binding site, which includes at least about two β-sheets, at least about two to about three short a-helixes, at least one small B-sheet and at least one additional small β- sandwich at the bottom of the molecule and at least about four disulfide bonds. The β-sandwich that includes the substrate binding site of the HAl-1 includes at least about four β-strands as the top sheet and at least about three to about four β-strands as the bottom sheet. At least about one a-helix of the HAl-1 portion is located by the side of β-sandwich that includes the substrate binding site and at least about one to about two are located at the bottom of the β-sandwich that includes the substrate binding site. The small β-sandwich of the HAl-1 can include at least about two to about three β-strands in each β-sheet; or about three to about four β-strands.

Exemplary HAl-1 protein portions include SEQ ID NOs: 8-44. In a particular embodiment, the HAl-1 portion of HA includes a portion of the globular head of HA that has a β-sheet and a β-sandwich at the bottom of the globular head that essentially retains the globular head in its tertiary structure. [00109] "HA1-2," as used herein, refers to a protein portion of a viral

hemagglutinin that includes at least about one B-sandwich that includes the substrate binding site, at least about two to about three short a-helixes, at least about one small B-sheet at the bottom of the molecule and at least about two disulfide bonds. A B- strand in a viral hemagglutinin can include between about two to about 15 amino acids. A small B-strand can include about two amino acids; or between about two to about three amino acids; or between about two to four amino acids or between about two to about five amino acids. A small B-sheet can include between about two to about three B-strands; or between about three to about four B-strands. The B- sandwich that includes the substrate binding site of HAI-2 can further include at least about four B-strands as the top sheet and at least about three to about four B- strands as the bottom sheet. At least about one a-helix of the HAI-2 portion is located by the side of the B-sandwich that includes the substrate binding site and at least about one to about two are located at the bottom of the B-sandwich that includes the substrate binding site. Exemplary HAI-2 protein portions include SEQ ID NOs: 45-97 and 287. In a particular embodiment, the HAI-2 portion of HA includes a B-sheet at the bottom of a portion of the globular head of hemagglutinin.

[00110] In another embodiment, the portion of an influenza viral hemagglutinin fused to at least one loop of domain 3 of at least a portion or the entirety of flagellin includes a portion of the globular head of hemagglutinin having a substrate binding site having at least one β-sheet that includes at least two β-sheets and comprises at least one β-sheet at a bottom of the portion of the globular head that causes the portion of the globular head to essentially retain its tertiary structure, at least about two to about three a-helices and at least about 4, at least about 5, at least about 6, at least about 7 or at least about 8 disulfide bonds and, optionally, at least one a-helix located by a side of the β-sandwich of the substrate binding site, and, optionally, at least one to two a-helices located at a bottom of the β-sandwich of the substrate binding site, and, optionally, the β-sheet at the bottom of the globular head includes between two to three β-strands or between three to four β-strands, and, optionally, the β-sandwich at the bottom of the globular head includes at least two to three β- strands in each β-sheet of the β-sandwich or at least three to four β-strands in each β- sheet of the β-sandwich, and, optionally, the β-sandwich of the substrate binding site includes at least four β-strands in a top sheet of the β-sandwich and at least three to four β-strands in a bottom sheet of the β-sandwich; and, optionally, lacks a membrane fusion domain; and lacks a transmembrane domain and lacks a cytoplasmic domain.

[00111] In yet another embodiment, the portion of an influenza viral

hemagglutinin fused to at least one loop of domain 3 of at least a portion or the entirety of flagellin includes a portion of the globular head of hemagglutinin having a substrate binding site that includes at least about one β-sandwich and comprises at least one β-sheet at a bottom of the portion of the globular head that causes the portion of the globular head to essentially retain its tertiary structure, and further includes at least about two to about three a-helices and at least about 2 disulfide bonds (or at least about 3 disulfide bonds, or at least about 4 disulfide bonds, or at least about 5 disulfide bonds, or at least about 6 disulfide bonds, or at least about 7 disulfide bonds), and, optionally, at least one a-helix located by a side of the β- sandwich of the substrate binding site, and, optionally, at least one to two a-helices located at a bottom of the β-sandwich of the substrate binding site; and, optionally, the β-sandwich of the substrate binding site includes at least four β-strands as a top sheet in the β-sandwich and at least three to four β-strands as a bottom sheet in the β-sandwich, and, optionally, lacks a membrane fusion domain; and lacks a transmembrane domain and lacks a cytoplasmic domain.

[00112] A single polypeptide can exhibit several types of secondary structure. Without any stabilizing interactions, a polypeptide can assume a random-coil conformation. However, secondary structures, such as alpha(a)-helices and beta(B)- strands, can stabilize a protein or a portion of a protein. Lateral association of β- strands form β-sheets (also referred to herein as "β-pleated sheets"). Secondary structures can be located at the surfaces of the antigen (e.g., portion of the viral hemagglutinin). A tertiary structure of a protein is the three-dimensional arrangement of amino acid residues. In contrast to secondary structure, which is stabilized by, for example, hydrogen bonds, a-helices, B-strands, tertiary structure results from hydrophobic interactions between non-polar side chains of the antigen, such as a portion of an influenza viral hemagglutinin. The hydrophobic interactions hold the helices strands in random coils in a compact internal scaffold. The size and shape of a protein can depend on its primary amino acid sequence, as well as the number, size and arrangement of secondary structures.

[00113] Portions of influenza viral hemagglutinins for fusion to at least one loop of domain 3 of flagellin include at least a portion of a globular head having secondary structures at the bottom of the globular head that essentially retains the globular head in its tertiary structure. Such secondary structures include at least one β-sheet; at least one β-sheet and at least one β-sandwich; at least one β-sheet, at least one β-sandwich and at least two β-strands.

[00114] Portions of hemagglutinin employed in the composition of the invention that are fused to at least one loop of domain 3 of flagellin can lack a transmembrane domain and a cytoplasmic domain. The portions of hemagglutinin employed in the compositions of the invention can further lack at least a portion of the HA2 subunit or the entirety of the HA2 subunit of hemagglutinin.

[00115] Antigens that have an isoelectric point greater than about 7.0 can be fused to at least one loop of domain 3 of flagellin to form fusion proteins of the invention. The loops in domain 3, which can be characterized as "extended regions," may accommodate amino acid insertions of about 50 amino acids, about 100 amino acids, about 150 amino acids, about 200 amino acids, about 250 amino acids, about 300 amino acids, about 350 amino acids, about 400 amino acids, about 450 amino acids and about 500 amino acids in length, including insertions between about 50 to about 100 amino acids, between about 100 to about 250 amino acids and between about 250 amino acids to about 500 amino acids with minimal disruption of the tertiary structure of flagellin. For example, the fusion protein of SEQ ID NO: 126 (HL772) includes a portion of an influenza B/WI1 that is 249 amino acids in length; the fusion protein of SEQ ID NO: 128 (HL657) includes a portion of an influenza B/BR60 that is 250 amino acids in length; the fusion protein of SEQ ID NO: 129 (HL775) includes a portion of an influenza A/PE16 that is 275 amino acids in length; and the fusion protein of SEQ ID NO: 130 (HL1018) includes a portion of an influenza A/AH1 that is 282 amino acids in length. [00116] The antigen with an isoelectric point greater than about 7.0 includes secondary structures, such as at least one β-sheet, at least one β-sandwich, at least one β-pleat and at least one a-helix, that maintain the antigen in its tertiary, compact configuration when the antigen is fused to a loop of domain 3 of flagellin, thereby minimally disrupting the tertiary structure of the loop of domain 3.

[00117] There are at least seventeen different HA antigens, the different HA antigens are classified as subtypes and identified as HA1 through HI 7. HI, H2 and H3 are found in human influenza viral antigens. Hemagglutinin on the influenza virion is a trimer (three copies of the HA polypeptides). The cleavage site for cell proteases on the HA protein is located near the viral membrane. The uncleaved form of hemagglutinin is referred to as HA0. After cleavage by a cellular enzyme, two subunits of HA are produced, specifically the HA1 subunit and the HA2 subunit. The two subunits remain together on the surface of the virus particle. The HA2 subunit that is produced by cleavage contains a sequence of hydrophobic acids referred to as a fusion peptide.

[00118] In an additional embodiment, the portion of hemagglutinin that is fused to at least one loop of domain 3 of flagellin has at least one β-sheet at a bottom of the portion of the globular head, such as HAl-2 portions of SEQ ID NOS: 45-97. In addition to at least one B-sheet at the bottom of a portion of the globular head, the portion of hemagglutinin fused to at least one loop of domain 3 of flagellin can further include at least one B-sandwich at the bottom of the globular head and, optionally, at least two B-strands at the bottom of the portion of the globular head. Thus, the portion of influenza viral hemagglutinin fused to at least one loop of domain 3 of flagellin can include at least one B-sheet, at least one B-sandwich, and at least two B-strands at the bottom of a portion of a globular head of influenza viral hemagglutinin, which is referred to herein a "HA1 -1L" portion of hemagglutinin. Exemplary HA1-1L portions of HA are SEQ ID NOS: 98-121, 228 and 273-277.

[00119] Fusion proteins of the invention can include an amino acid linker between at least one of an amino-terminus or a carboxy-terminus of the antigen and the loop of domain 3 of the flagellin. The amino acid linker can be between about 1 to about 10 amino acids in length, such as about 2, about 3, about 4, about 5, about 8 or about 9 amino acids in length. Preferred amino acid residues would include amino acid residues without side chains or amino acid residues with small side chains, such as glycine, alanine or serine, including combinations of glycine, serine and alanine. For example, amino acid residues of a linker of at least about 9 amino acids in length could be fused to the carboxy-terminus of an antigen of SEQ ID NO: 125 (HL352) in a fusion protein. Exemplary antigens fused to a linker are SEQ ID NOS: 46, 48, 50, 52, 54, 56, 58, 60, 62, 89, 91, 93-97, 228, 273-277. Amino acid residues that comprise the amino acid linker can have an isoelectric point of between about 3 and about 7 and between about 4 and about 7. Amino acid linkers can include 2, 3, 4 or 5 negatively charged amino acid residues, such as aspartic acid or glutamic acid. Exemplary amino acid linkers are described, for example, in

PCT/US2012/000099 (WO 2012/115715), by Song, L. et al, and

PCT/US2012/000367 (WO 2013/066365), by Song, L., et al. It is believed that negatively charged amino acids at or adjacent to (about 1, 2, 3 or 4 amino acids) the amino- or carboxy-terminus of the antigen at the site of fusion to a loop of domain 3 reduce undesirable intramolecular interactions between the negatively charged flagellin and a positively charged antigen thereby tethering the antigen at the site of fusion. Amino acid linkers are devoid of secondary structures, such as a-helices and β-sheets.

[00120] In an embodiment, the linker can include amino acid residues that are native to the naturally occurring hemagglutinin and adjacent to the portion of HA (e.g., HAl-1, HA1-2, HA1-1L) being fused to a loop of domain 3, such as the 9 amino acid residues (also referred to herein as "extension") of SEQ ID NO: 123, which includes 2 negatively charged amino acid residues. For example, the fusion protein of SEQ ID NO: 154 (HL719, which is an R3 construct) includes a portion of influenza B Wisconsin construct (HA1-2 of SEQ ID NO: 50) and 9 amino acids (SEQ ID NO: 123) of the naturally occurring hemagglutinin, which has 2 negatively charged residues. Exemplary antigens fused to the same linker are SEQ ID NOS: 46, 48, 52, 54, 56, 58, 60, 62, 89, 91, and 93. The fusion protein of SEQ ID NO: 152 (HL724, which is an R3 construct) includes a portion of an influenza B

Wisconsin HA 1-2 that has the 9 amino acids of SEQ ID NO: 123 with an additional 3 negatively charged amino acid residues in the portion of HA. Exemplary antigens fused with the same modifications are SEQ ID NOS: 94-97.

[00121] An amino acid linker used to fuse the antigen to at least one loop of domain 3 of flagellin can be about 1 to about 10 amino acids in length, such as about 2 amino acids, about 3 amino acids, about 4 amino acids, about 5 amino acids, about 6 amino acids, about 7 amino acids, about 8 amino acids, about 9 amino acids and about 10 amino acids in length. Preferably, amino acid linkers include amino acid residues that have small or no side chains, such as glycine, alanine and serine. In a particular embodiment, the amino acid linker that would be fused to the carboxy- terminus of the loop of domain 3 of flagellin has an overall negative charge, such as an isoelectric point between about 2 to about 4.

[00122] In a further embodiment, the invention is a composition comprising at least three fusion proteins that each activate a Toll-like Receptor 5. The first fusion protein includes a first flagellin and a first influenza A viral hemagglutinin antigen that has an isoelectric point greater than about 6.0, such as an isoelectric point between about 6.5 to about 7.0, fused to a portion of the first flagellin. The second fusion protein includes a second flagellin and a second influenza A viral

hemagglutinin antigen that is distinct from the first influenza A viral hemagglutinin antigen and that has an isoelectric point greater than about 7.0, such as an isoelectric point between about 7.5 to about 8.5, about 7.8, about 8.0, about 8.5, about 9.0 and about 9.5, fused to a portion of the second flagellin. The third fusion protein includes a third flagellin and a first influenza B viral hemagglutinin antigen that has an isolectric point greater than about 8.0, such as between about 8.0 to about 10.0, about 8.5, about 9.0, about 9.5 and about 10.0, and is fused to at least one loop of domain 3 of the third flagellin.

[00123] The composition that includes at least three fusion proteins that activate TLR5 can further include a fourth fusion protein that activates TLR 5. The fourth fusion protein includes a second influenza B viral hemagglutinin antigen that is distinct from the first influenza B antigen, and, that has an isolectric point greater than about 8.0, , such as between about 8.0 to about 10.0, about 8.5, about 9.0, about 9.5 and about 10.0, fused to at least one loop of domain 3 of a fourth flagellin. [00124] In a particular embodiment, the first fusion protein of the composition that includes at least three fusion proteins that each activate TLR5 is a fusion protein that includes two similar portions of an influenza A hemagglutinin that have a portion of a globular head with a β-sheet at the bottom of the globular head (e.g., HAl -2) fused to an R32x construct of flagellin. This first fusion protein has one HA portion fused to the portion of the flagellin in the region where domain 3 was located and another similar HA portion fused to the carboxy-terminus of the portion of flagellin. An exemplary first fusion protein is SEQ ID NO: 268, which is a portion of influenza A California/07/2009.

[00125] In a particular embodiment, the second fusion protein of the composition that includes at least three fusion proteins that each activate TLR5 is a fusion protein that includes a second influenza A antigen that is distinct from the first influenza A antigen and has an isoelectric point greater than about 7.0. The second influenza A antigen is a portion of the globular head of HA that has a β-sheet, a one β-sandwich and two B-strands at the bottom of a portion of a globular head of influenza viral hemagglutinin (referred to herein a "HA1-1L") fused to an R3 construct of a portion of flagellin. The HA1-1L antigen replaces domain 3 of the flagellin in an R3 construct. An exemplary second fusion protein is SEQ ID NO: 269, which is a portion of influenza A Perth/16/2009.

[00126] In a particular embodiment, the third fusion protein of the composition that includes at least three fusion proteins that each activate TLR5 is a fusion protein that includes a first influenza B antigen that has an isoelectric point greater than about 8.0 and that is fused to a loop of domain 3 of a portion of a third flagellin, such as loop 6 between amino acid residues 277 and 278 of SEQ ID NO: 2. An exemplary third fusion protein is SEQ ID NO: 126 (HL772), which is a portion of influenza B Wisconsin/1/2010.

[00127] In a particular embodiment, the fourth fusion protein of the composition that includes at least three fusion proteins that each activate TLR5 is a fusion protein that includes a second influenza B antigen that has an isoelectric point greater than about 8.0 and that is fused to a loop of domain 3 of a portion of a fourth flagellin, such as loop 6 between amino acid residues 277 and 278 of SEQ ID NO: 2. The second influenza B antigen is distinct from the first influenza B antigen. An exemplary fourth fusion protein is SEQ ID NO: 158 (HL787), which is a portion of influenza B Bangladesh/5495/2009.

[00128] "Distinct," as used herein in reference to an antigen, protein or flagellin, means that the antigen, protein or flagellin is different than a first or additional (second, third, fourth, etc.) antigen, first or additional (second, third, fourth, etc.) protein or first or additional (second, third, fourth, etc.) flagellin. For example, an influenza A Perth/16/2009 hemagglutinin (HA) antigen is distinct from an influenza A California/07/2009 HA antigen. Likewise, an influenza B Wisconsin/ 1/2010 HA antigen is disctinct from an influenza B Bangladesh/5495/2009 HA antigen.

[00129] FIG 31 depicts the domains (DO, Dl, D2, D3) of a flagellin construct (SEQ ID NO: 283) and a fusion protein that includes, in sequence, the amino- domain 0, the amino-domain 1, the amino-domain 2, the carboxy-domain 2, the carboxy-domain 1 and the carboxy-domain 0 of flagellin. An antigen (Ag) is fused between the amino- and carboxy-domain 2 of the flagellin construct. The flagellin construct of the fusion protein depicted in FIG. 31 lacks the D3 domain of flagellin and is referred to herein as an "R3 construct" or the "R3 form of flagellin" or "R3 flagellin construct" (SEQ ID NO: 283). "R3 (Replace Domain 3) construct," as used herein, means that Domain 3 of the flagellin has been replaced with an antigen, as previously described in U.S. Patent Application No: 12/905,584. For example, an antigen can be fused between amino acids 190 and 191 of SEQ ID NO: 283. >

[00130] FIG. 32 depicts the domains (DO, Dl, D2, D3) of a flagellin construct (SEQ ID NO: 284) and a fusion protein that includes, in sequence, the amino- domain 0, the amino-domain 1 , the amino-domain 2, the carboxy-domain 2, the carboxy-domain 1 and the carboxy-domain 0 of flagellin. For example, the fusion protein depicted in SEQ ID NO: 284 has at least a portion of at least one antigen and at least a portion of at least one other antigen fused to two sites in a portion of flagellin. One antigen (also referred to as "first antigen") is fused between the amino-domain 2 and the carboxy-domain 2 of the flagellin construct. The other antigen, which can be distinct from the first antigen, is fused to the carboxy-terminal amino acid of the domain 0 of the flagellin construct. The flagellin construct depicted in FIG. 32 lacks the D3 domain of flagellin and is referred to herein as an "R3-2xAg construct" or the "R3-2xAg form of flagellin" or "R32x flagellin construct" or "R32x" or the "R32x form of flagellin" or "2xR3" or the "2xR3 form of flagellin" or the "R3/2x form of flagellin." "R3-2xAg construct," as used herein, means that Domain 3 of the flagellin is replaced with a one antigen and another antigen, such as a distinct antigen, is fused to the carboxy-terminus of domain 0, as previously described in U.S. Patent Application No: 12/905,584. For example, one antigen can be fused between amino acids 190 and 191 of SEQ ID NO: 284 and another antigen can be fused to the terminal carboxy amino acid (e.g., SEQ ID NOs: 268, 287 and 288).

[00131] In still another embodiment, the first influenza A viral hemagglutinin antigen of the compositions of the invention and for use in the methods of the invention includes at least a portion or the entirety of an HAl subunit that has at least a portion of a globular head that includes at least one β-sheet at a bottom of the globular head; the second influenza A viral hemagglutinin antigen of the

compositions of the invention and for use in the methods of the invention includes at least a portion or the entirety of an HAl subunit having at least a portion of a globular head that includes at least one β-sheet, at least one β-sandwich and at least two β-strands at the bottom of the portion of the globular head; and the first influenza B viral hemagglutinin antigen of the compositions of the invention and for use in the methods of the invention includes at least a portion or the entirety of an HAl subunit that has at least a portion of a globular head that includes at least one β-sheet at a bottom of the globular head. The second influenza B viral

hemagglutinin antigen of the compositions of the invention and for use in the methods of the invention includes at least a portion or the entirety of an HAl subunit that has at least a portion of a globular head that includes at least one β-sheet at a bottom of the globular head.

[00132] In a further embodiment, the first influenza A viral hemagglutinin antigen of the compositions of the invention and for use in the methods of the invention includes at least a portion of an HAl subunit that has at least a portion of a globular head that includes at least one β-sheet at a bottom of the globular head; the second influenza A viral hemagglutinin antigen of the compositions of the invention and for use in the methods of the invention includes at least a portion of an HA1 subunit having at least a portion of a globular head that includes at least one β-sheet, at least one β-sandwich and at least two β-strands at the bottom of the portion of the globular head; and the first influenza B viral hemagglutinin antigen of the compositions of the invention and for use in the methods of the invention includes at least a portion of an HA1 subunit that has at least a portion of a globular head that includes at least one β-sheet at a bottom of the globular head. The second influenza B viral hemagglutinin antigen of the compositions of the invention and for use in the methods of the invention includes at least a portion of an HA1 subunit that has at least a portion of a globular head that includes at least one β-sheet at a bottom of the globular head.

[00133] In another embodiment, the invention is a method of stimulating an immune response to an antigen in a subject. The method includes the step of administering to the subject a composition that includes a fusion protein that activates a Toll-like Receptor 5, the fusion protein comprising a flagellin and at least one antigen that has an isoelectric point greater than about 7.0 and that is fused to at least one loop of domain 3 of the flagellin. In a particular embodiment, the immune response that is stimulated in the subject in response to administration of the fusion protein provides protective immunity to a disease or an infection consequent to exposure of an organism that includes the antigens fused to the loop of domain 3 of flagellin. The method can further include the administration of an adjuvant.

[00134] In yet another embodiment, the invention is a method of stimulating an immune response in a subject, comprising the step of administering to the subject a composition that includes at least three fusion proteins each of which activates a Toll-like Receptor 5, wherein: (a) a first fusion protein includes a first flagellin and a first influenza A viral hemagglutinin antigen that has an isoelectric point greater than about 6.0 fused to a portion of the first flagellin; (b) a second fusion protein includes a second flagellin and a second influenza A viral hemagglutinin antigen that is distinct from the first influenza A viral hemagglutinin antigen and that has an isoelectric point greater than about 7.0 fused to a portion of the second flagellin; and (c) a third fusion protein includes a third flagellin and a first influenza B viral hemagglutinin antigen that has an isolectric point greater than about 8.0 and is fused to at least one loop of domain 3 of the third flagellin.

[00135] In a further embodiment, the invention is a method of stimulating an immune response in a subject, comprising the step of administering to the subject a composition that includes at least four fusion proteins each of which activates a Tolllike Receptor 5, wherein: (a) a first fusion protein includes a first flagellin and a first influenza A viral hemagglutinin antigen that has an isoelectric point greater than about 6.0 fused to a portion of the first flagellin; (b) a second fusion protein includes a second flagellin and a second influenza A viral hemagglutinin antigen that is distinct from the first influenza A viral hemagglutinin antigen and that has an isoelectric point greater than about 7.0 fused to a portion of the second flagellin; (c) a third fusion protein includes a third flagellin and a first influenza B viral hemagglutinin antigen that has an isolectric point greater than about 8.0 and is fused to at least one loop of domain 3 of the third flagellin; and (d) a fourth fusion protein that activates a Toll-like Receptor 5 includes a second influenza B viral

hemagglutinin antigen that is distinct from the first influenza B viral hemagglutinin antigen and that has an isoelectric point greater than about 8.0 fused to at least one loop of domain 3 in the fourth flagellin.

[00136] In an embodiment, the method of the invention stimulates an immune response by administering four different fusion proteins (2 fusion proteins that include influenza A antigens and two fusion proteins that include influenza B antigens) each of which activate TLR5 signaling. The four fusion proteins can be administered in escalating doses. For example, an initial immunization can include all four fusion proteins administered in the same dose (e.g., about 1 μg or about 2 μg), followed by a second dose that is twice the first dose (e.g., about 2 μg or about 4 μg). Subsequent doses (a third or a fourth dose) can include further increases in the doses of fusion proteins that include influenza A to doses of about 6 μg to about 8 μg, while maintaining the dose of the fusion protein that includes influenza B at about 2 μg or 4 μg. [00137] Doses of fusion proteins that include influenza A antigens in multivalent compositions can be at least one member selected from the group consisting of about 2 g, about 4 μg, about 6 μg, about 10 ^g, about 12 μg, about 14 μg, about 15 μg and about 20 μg doses. Doses of fusion proteins that include influenza B antigens in multivalent compositions can be at least one member selected from the group consisting of about 1 μg, about 2 μg, about 4 μg, about 6 μg, about 8 μg and about 10 μg doses. Multivalent compositions include compositions of the invention that have at least three fusion proteins, such as a composition that includes two fusion proteins having influenza A antigens and one or two fusion proteins having influenza B antigens. The fusion protein of the composition administered to the subject can be in a dose of at least one member selected from the group consisting of about 1 μg, about 2 μg, about 3 μg, about 4 μg, about 5 μg, about 6 μg, about 7 μg, about 8 μg, about 9 μg, about 10 μg dose, about 15 μg dose, about 20 μg dose, about 25 μg dose, about 30 μg dose, about 35 μg dose, about 40 μg dose, about 45 μg dose and about 50 μg dose.

[00138] In still another embodiment, the invention is a method of stimulating an immune response to influenza A and influenza B antigens in a subject. The method includes the step of administering to the subject a composition that includes SEQ ID NOs: 268, 269, 126, and 158.

[00139] The methods of the invention can further include the administration of an adjuvant.

[00140] "Stimulating an immune response," as used herein, refers to the generation of antibodies and/or T-cells to the antigen fused to a loop of domain 3 of flagellin, such as the protein portions of influenza B hemagglutinin (HA) (e.g., HA1- 1, HA1-2, HA1-1L proteins) described herein. Stimulating an immune response in a subject can include the production of humoral and/or cellular immune responses that are reactive against the antigen, such as a viral protein, in particular, an influenza viral protein.

[00141] The compositions of the invention for use in methods to stimulate immune responses in subjects, can be evaluated for the ability to stimulate an immune response in a subject using well-established methods. Exemplary methods to determine whether the compositions of the invention stimulate an immune response in a subject, include measuring the production of antibodies specific to the antigen (e.g., IgG antibodies) by a suitable technique such as, ELISA assays; the potential to induce antibody-dependent enhancement (ADE) of a secondary infection; macrophage-like assays; neutralization assessed by using the Plaque Reduction Neutralization Test (PRNTgo); and the ability to generate serum antibodies in non-human models (e.g., mice, rabbits, monkeys) (Putnak, et al, Vaccine 25:4442-4452 (2005)).

[00142] "Stimulates a protective immune response," as used herein, means administration of the compositions of the invention that include a fusion protein comprising, for example, an influenza B antigen fused to at least one loop of domain 3 of flagellin that result in production of antibodies to the protein to thereby cause a subject to survive challenge by an otherwise lethal dose of a viral protein, such as influenza B viral challenge. Techniques to determine a lethal dose of a virus (e.g., an influenza B virus) are known to one of skill in the art (see, for example,

WHO/CDS/CSR/NCS2002.5 "WHO Manual on Animal Influenza Diagnosis and Surveillance" World Health Organization, Dept of Communicable Disease

Surveillance and Response, WHO Global Influenza Programme; Harmon, M.W., et al, J. Clin. Microbiol. 2(5:333-337 (1988); Reed, L.J., et al, Am. J. Hyg. 27:493-497 (1938); Rose, T., et al, J. Clin. Microbiol. 37:937-943 (1999); Walls, H.H. et al, J. Clin. Microbiol. 23:240-245 (1986); Current Protocols in Immunology, 19.1 1.1- 19.11.32, Cottey, R., et al, John Wiley & Sons, Inc (2001)). Exemplary techniques for determining a lethal dose can include administration of varying doses of virus and a determination of the percent of subjects that survive following administration of the dose of virus (e.g., LD10, LD₂₀, LD₄₀, LD50, LD₆₀, LD₇₀, LDgo, LD90). For example, a lethal dose of a virus that results in the death of 50% of a population of subjects is referred to as an "LD₅0"; a lethal dose of a virus that results in the death of 80% of a population of subjects is referred to herein as "LD₈₀"; a lethal dose of a virus that results in death of 90% of a population of subjects is referred to herein as "LD₉o." [00143] For example, determination of the LD90 can be conducted in subjects (e.g., mice) by administering intranasally varying doses (e.g., dilutions, such as log and half-log dilutions of 8 x 10³ egg-infectious doses (EID)) followed by an assessment of the survival of the subjects about 14 days to about 21 days after infection with the virus. Protective immunity can be assessed by physical appearance of the subject, general demeanor (active), weight (initial loss of weight followed by return to a weight about the weight of the subject prior to infection with the virus) and survival after about 14 to about 21 days following infection with the virus.

[00144] Assessment of stimulation of protective immunity can also be made by employing assays that assess the ability of the antibodies produced in response to the compositions of the invention (e.g., an antigen, such as an influenza B viral antigen, such as a portion of hemagglutinin) to neutralize binding of the viral protein (e.g., hemagglutinin protein) to a host cell (see, for example, Current Protocols in

Immunonology, 19.11.1-19.11.32, Cottey, R., et al, John Wiley & Sons, Inc (2001)). Assessment of stimulation of protective immunity can also be made by employing assays that measure the ability of antibodies to inhibit hemagglutinin binding (see, for example, Burnett, F.M., et al, J. exp. Biol. Med. Sci. 25:227-233 (1947); Salk, J.E. J. Immunol. 49:87-98 (1944); Current Protocols in Immunology, 19.11.1-19.11.32, Cottey, R., et al., John Wiley & Sons, Inc (2001)).

[00145] It is believed that inhibition of hemagglutinin binding is indicative of the ability of antibodies, formed from the compositions and by the methods of the invention, to neutralize the sialic acid binding sites of the naturally occurring viral hemagglutinin ("neutralization of HA binding") and, thereby, prevent infection of the host cell as a consequence of stimulating a protective immune response.

Inhibition or neutralization of hemagglutinin binding is believed to correlate with an ability of an immune response to protect against a lethal dose of virus.

[00146] Neutralization of HA binding can be assessed by in vitro assays (See, for example, Current Protocols in Immunology 19.11.1-19.11.32, Cottey, R., et al, Suppl. 42, John Wiley & Sons, Inc. (2001) and WHO Manual on Animal Influenza Diagnosis and Surveillance, Webster, R., et al., pages 28-36, 48-54, 82-92 (2002)). Exemplary viral neutralization assays rely on the ability of serum to specifically bind and prevent replication of influenza virus in culture, such as in the Madin- Darby Canine Kidney (MDCK) cell line. Briefly, cells are cultured in 96 well plates in the presence of a previously titered virus and the cytopathic effect of the replicating virus is observed under a microscope. To test serum, serial dilutions of the serum are prepared and preincubated with the viral stock for about 2 hours at 37°C prior to infecting the MDCK cells. The mixture is incubated for an additional 2 hours after which the virus/serum mixture is removed and replaced with fresh media. The cells are grown for 4 days. Wells are scored as positive for viral growth if at least about 50% of the cells are dead in at least about half of the wells for a given serum dilution. The reciprocal of the highest dilution of serum which protects at least about half of the cells from death, in at least about half of the wells, is considered the neutralization titer.

[00147] Alternatively, a micro -neutralization in vitro assay can be performed to assess neutralization of HA binding. For example, serum is diluted and

preincubated with a known titer of virus and mixed with MDCK cells, as described above. After 2 days of incubation, cells are washed and fixed with acetone. The plates are developed as an ELISA using a monoclonal antibody to the influenza nuclear antigen NP. A microneutralization titer is determined as the reciprocal of the highest dilution which yields less than about 50% of the anti-NP reading of the virus-only control wells.

[00148] The Hemagglutination Inhibition (HAI) assay is based on the HA antigen on the surface of the influenza virus agglutinating red blood cells (RBC) and preventing red blood cells from precipitating. Antibodies that specifically bind the sialic acid-binding regions of HA prevent agglutination allowing precipitation. The assay is performed in 96 well V bottom plates with fresh chicken RBC. A stock of viral antigen is titered so that about a 4-fold excess of antigen is present relative to the minimum amount needed to prevent precipitation. The test serum, which can be from several species including mouse, ferret, poultry or human, is heated to about 56°C to inactivate complement. Serial 2-fold dilutions of the inactivated serum are performed and mixed with the stock HA. After about 30 minutes at room temperature, the RBCs are added and the plate is incubated for about 30 to about 45 minutes. Results are scored by observations: agglutination results in cloudy wells while inhibition results in a "button" of red cells precipitated at the bottom of the well. Controls include RBC with no HA, which forms a button, and HA and RBC with no serum, which remains cloudy. The HAI titer of a particular serum sample is the reciprocal of the last dilution which prevents agglutination (i.e., forms a button). For example, if about a 1 : 128 dilution reads as a button but the 1 :256 dilution does not, the HAI titer is about 128.

[00149] Fusion proteins of the invention can be made employing routine molecular biological techniques, as described herein. Host cells can be transfected with nucleic acids encoding fusion proteins of the invention. The host cells can be eukaryotic or prokaryotic host cells. Suitable prokaryotic host cells include E. coli, B. subtilis and Pseudomonas fluorescens.

[00150] The eukaryotic host cells employed in the methods of the invention can include a Saccharomyces eukaryotic host cell, an insect eukaryotic host cell (e.g., at least one member selected from the group consisting of a Baculovirus infected insect cell, such as Spodoptera frugiperda (Sf9) or Trichhoplusia ni (High5) cells; and a Drosophila insect cell, such as Dmel2 cells), a fungal eukaryotic host cell, a parasite eukaryotic host cell (e.g., a Leishmania tarentolae eukaryotic host cell), CHO cells, yeast cells (e.g., Pichia) and a Kluyveromyces lactis host cell.

[00151] Suitable eukaryotic host cells and vectors can also include plant cells (e.g., tomato; chloroplast; mono- and dicotyledonous plant cells; Arabidopsis thaliana; Hordeum vulgare; Zea mays; potato, such as Solarium tuberosum; carrot, such as Daucus carota L. ; and tobacco, such as Nicotiana tabacum, Nicotiana benthamiana (Gils, M., et al, Plant Biotechnol J. 5:613-20 (2005); He, D.M., et al, Colloids SurfB Biointerfaces, (2006); Huang, Z., et al, Vaccine 79:2163-71 (2001); Khandelwal, A., et al, Virology. 305:207-15 (2003); Marquet-Blouin, E., et al, Plant Mol Biol 57:459-69 (2003); Sudarshana, M.R., et al. Plant Biotechnol J.

:551-9 (2006); Varsani, A., et al, Virus Res, 120:91-6 (2006); Kamarajugadda S., et al, Expert Rev Vaccines 5:839-49 (2006); oya V, et al, Infect Immun. 75:8266- 74 (2005); Zhang, X., et al, Plant Biotechnol J. 4:419-32 (2006)). [00152] The fusion proteins of the invention can be purified and characterized employing well-known methods (e.g., gel chromatography, cation exchange chromatography, SDS-PAGE), as described herein.

[00153] For large scale production, fermentation techniques can be employed. Exemplary fermentation techniques can include a proposed cycle that can start with a culture inoculated into 6 L of MRBR media, as described herein, held at about 30°C, about pH 7, and DO controlled to greater than about 30%. A 6 liter feed can then be started at least about 30 minutes after glucose exhaustion. The proposed 6 liter feed media, when combined with 6 L of MRBR media, can provide the necessary conditions for E. coli growth based on about 52%> utilization of carbon for growth. The feed may or may not include IPTG. The batch can be induced with at least 2 mM IPTG, introduced as a bolus, shortly after the feed is started to initiate production. The feed rate can start at about 20 mL feed per hour per liter bioreactor volume and increase over time based on the ability of the culture to accept more glucose without glucose accumulation. The culture can be harvested when the feed is complete. The 6 liter feed media, about pH 6.0, can include Glucose 180 g/L; KH₂P0₄2 g/L; Na¾P0₄ (H₂0) 4 g/L; (NH₄)₂HP0₄ 12 g/L; (NH₄)₂HS0₄ 4 g/L; DL- Alanine 40 g/L; Citric Acid 4 g/L; MgSO₄(7H₂0) 5.5 g/L; Trace Metals 6 mL; CaCl₂ 2.5 g/L; FeS0₄ 7¾0 1 g/L.

[00154] Cell disruption and clarification in a large scale production can include removal of Triton X-100 from the resuspension buffer; dissolution of insolubles by the addition of 50 mM Tris, 25 mM NaCl, 8 M urea, about pH 8 to the lysate;

addition of PEI (polyethylamine) and subsequent removal by centrifugation with one or more of the buffers to remove nucleic acids and/or aid in filtration; the addition of flocullants, such as AEROSIL 380, AEROSIL 200, ALKOXIDE ALU C, and CELPUR; and subsequent removal by centrifugation to aid in filtration. Cation exchange chromatography can include the use of a process resin, adding a denaturing endotoxin removal step containing up to 8 M urea and up to about 2% Triton X-100, and a step gradient elution. The step elution gradient can include about 100 to about 200 mM NaCl. [00155] In an additional embodiment, the invention includes a protein, polypeptide or peptide having at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 98% and at least about 99% sequence identity to the proteins, polypeptides and peptides of the invention.

[00156] The percent identity of two amino acid sequences (or two nucleic acid sequences) can be determined by aligning the sequences for optimal comparison purposes (e.g., gaps can be introduced in the sequence of a first sequence). The amino acid sequence or nucleic acid sequences at corresponding positions are then compared, and the percent identity between the two sequences is a function of the number of identical positions shared by the sequences (i.e., % identity = # of identical positions/total # of positions x 100). The length of the protein or nucleic acid encoding can be aligned for comparison purposes is at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 98%, at least about 99% of the length of the reference sequence, for example, the nucleic acid sequence of a flagellin, including a nucleic acid sequence that encodes an R3 construct (e.g., SEQ ID NO: 283) or an R32x construct (e.g., SEQ ID NO: 284, 285), flagellin (SEQ ID NOs: 2, 127), antigen (e.g., SEQ ID NOs: 50, 51, 52, 98 and 287) employed in the fusion proteins of the invention of a fusion protein of the invention and fusion proteins of the invention (e.g., SEQ ID NOs:126, 158, 268 and 269).

[00157] The actual comparison of the two sequences can be accomplished by well-known methods, for example, using a mathematical algorithm. A preferred, non-limiting example of such a mathematical algorithm is described in Karlin et al. (Proc. Natl. Acad. Sci. USA, 0:5873-5877 (1993), the teachings of which are hereby incorporated by reference in its entirety). Such an algorithm is incorporated into the BLASTN and BLASTX programs as described, for example, in Schaffer et al. {Nucleic Acids Res., 29:2994-3005 (2001) and Oehmen, C. S. et al,

Bioinformatics 29 (6): 797-798 (2013). When utilizing BLAST and Gapped BLAST programs, the default parameters of the respective programs (e.g.,

BLASTN; available at the Internet site for the National Center for Biotechnology Information) can be used. In one embodiment, the database searched is a non- redundant (NR) database, and parameters for sequence comparison can be set at: no filters; Expect value of 10; Word Size of 3; the Matrix is BLOSUM62; and Gap Costs have an Existence of 11 and an Extension of 1.

[00158] Another mathematical algorithm employed for the comparison of sequences is the algorithm of Myers and Miller, CABIOS (1989). Such an algorithm is incorporated into the ALIGN program (version 2.0), which is part of the GCG (Accelrys, San Diego, California) sequence alignment software package. When utilizing the ALIGN program for comparing amino acid sequences, a

PAM120 weight residue table, a gap length penalty of 12, and a gap penalty of 4 is used. Additional algorithms for sequence analysis are known in the art and include ADVANCE and ADAM as described in Torellis, et al, Comput, Appl. Biosci., 10: 3-5 (1994) and FASTA described in Pearson, et al, (Proc. Natl. Acad. Sci USA, 85: 2444-2448 (1988), the teachings of which are hereby incorporated by reference in its entirety).

[00159] The percent identity between two amino acid sequences can also be accomplished using the GAP program in the GCG software package (Accelrys, San Diego, California) using either a Blossom 63 matrix or a PAM250 matrix, and a gap weight of 12, 10, 8, 6, or 4 and a length weight of 2, 3, or 4. In yet another embodiment, the percent identity between two nucleic acid sequences can be accomplished using the GAP program in the GCG software package (Accelrys, San Diego, California), using a gap weight of 50 and a length weight of 3.

[00160] The nucleic acid sequence encoding a protein portion of HA, polypeptide or fusion proteins of the invention and polypeptides of the invention can include nucleic acid sequences that hybridize to nucleic acid sequences or complements of nucleic acid sequences of the invention, for example, the nucleic acid sequence of a flagellin, including an R3 construct or an R32x construct, antigen employed in the fusion proteins of the invention of a fusion protein of the invention under selective hybridization conditions (e.g., highly stringent hybridization conditions). As used herein, the terms "hybridizes under low stringency," "hybridizes under medium stringency," "hybridizes under high stringency," or "hybridizes under very high stringency conditions," describe conditions for hybridization and washing of the nucleic acid sequences. Guidance for performing hybridization reactions, which can include aqueous and nonaqueous methods, can be found in Aubusel, F.M., et al, Current Protocols in Molecular Biology, John Wiley & Sons, N.Y. (2001).

[00161] For applications that require high selectivity, relatively high stringency conditions to form hybrids can be employed. In solutions used for some membrane based hybridizations, addition of an organic solvent, such as formamide, allows the reaction to occur at a lower temperature. High stringency conditions are, for example, relatively low salt and/or high temperature conditions. High stringency are provided by about 0.02 M to about 0.10 M NaCl at temperatures of about 50°C to about 70°C. High stringency conditions allow for limited numbers of mismatches between the two sequences. In order to achieve less stringent conditions, the salt concentration may be increased and/or the temperature may be decreased. Medium stringency conditions are achieved at a salt concentration of about 0.1 to 0.25 M NaCl and a temperature of about 37°C to about 55°C, while low stringency conditions are achieved at a salt concentration of about 0.15 M to about 0.9 M NaCl, and a temperature ranging from about 20°C to about 55°C. Selection of components and conditions for hybridization are well known to those skilled in the art and are reviewed in Ausubel et al, Short Protocols in Molecular Biology, John Wiley & Sons, New York N.Y., Units 2.8-2.1 1 , 3.18-3.19 and 4-64.9, (1997).

[00162] A "subject," as used herein, can be a mammal, such as a primate or rodent (e.g., rat, mouse). In a particular embodiment, the subject is a human.

[00163] An "effective amount," when referring to the amount of a composition and fusion protein of the invention, refers to that amount or dose of the composition and fusion protein, that, when administered to the subject is an amount sufficient for therapeutic efficacy (e.g., an amount sufficient to stimulate an immune response in the subject). The compositions and fusion proteins of the invention can be administered in a single dose or in multiple doses.

[00164] The methods of the present invention can be accomplished by the administration of the compositions and fusion proteins of the invention by enteral or parenteral means. Specifically, the route of administration is by oral ingestion (e.g., drink, tablet, capsule form) or intramuscular injection of the composition and fusion protein. Other routes of administration as also encompassed by the present invention including intravenous, intradermal, intraarterial, intraperitoneal, or subcutaneous routes, and nasal administration. Suppositories or transdermal patches can also be employed.

[00165] The compositions and fusion proteins of the invention can be

administered ex vivo to a subject's autologous dendritic cells. Following exposure of the dendritic cells to the composition and fusion protein of the invention, the dendritic cells can be administered to the subject.

[00166] The compositions and fusion proteins of the invention can be

administered alone or can be coadministered to the patient. Coadminstration is meant to include simultaneous or sequential administration of the composition, fusion protein or polypeptide of the invention individually or in combination.

Where the composition and fusion protein are administered individually, the mode of administration can be conducted sufficiently close in time to each other (for example, administration of the composition close in time to administration of the fusion protein) so that the effects on stimulating an immune response in a subject are maximal. It is also envisioned that multiple routes of administration (e.g., intramuscular, oral, transdermal) can be used to administer the compositions and fusion proteins of the invention.

[00167] The compositions and fusion proteins of the invention can be

administered alone or as admixtures with conventional excipients, for example, pharmaceutically, or physiologically, acceptable organic, or inorganic carrier substances suitable for enteral or parenteral application which do not deleteriously react with the extract. Suitable pharmaceutically acceptable carriers include water, salt solutions (such as Ringer's solution), alcohols, oils, gelatins and carbohydrates such as lactose, amylose or starch, fatty acid esters, hydroxymethycellulose, and polyvinyl pyrolidine. Such preparations can be sterilized and, if desired, mixed with auxiliary agents such as lubricants, preservatives, stabilizers, wetting agents, emulsifiers, salts for influencing osmotic pressure, buffers, coloring, and/or aromatic substances and the like which do not deleteriously react with the compositions, fusion proteins or polypeptides of the invention. The preparations can also be combined, when desired, with other active substances to reduce metabolic degradation. The compositions and fusion proteins of the invention can be administered by is oral administration, such as a drink, intramuscular or

intraperitoneal injection or intranasal delivery. The compositions and fusion proteins alone, or when combined with an admixture, can be administered in a single or in more than one dose over a period of time to confer the desired effect (e.g., alleviate prevent viral infection, to alleviate symptoms of virus infection, such as influenza or flavi viral infection).

[00168] When parenteral application is needed or desired, particularly suitable admixtures for the compositions and fusion proteins are injectable, sterile solutions, preferably oily or aqueous solutions, as well as suspensions, emulsions, or implants, including suppositories. In particular, carriers for parenteral administration include aqueous solutions of dextrose, saline, pure water, ethanol, glycerol, propylene glycol, peanut oil, sesame oil, polyoxyethylene-block polymers, and the like.

Ampules are convenient unit dosages. The compositions, fusion proteins or polypeptides can also be incorporated into liposomes or administered via

transdermal pumps or patches. Pharmaceutical admixtures suitable for use in the present invention are well-known to those of skill in the art and are described, for example, in Pharmaceutical Sciences (17th Ed., Mack Pub. Co., Easton, PA) and WO 96/05309 the teachings of which are hereby incorporated by reference.

[00169] The compositions and fusion proteins of the invention can be

administered to a subject on a support that presents the compositions, fusion proteins and polypeptides of the invention to the immune system of the subject to generate an immune response in the subject. The presentation of the compositions, fusion proteins and polypeptides of the invention would preferably include exposure of antigenic portions of the viral protein to generate antibodies. The fusion proteins of the invention are in close physical proximity to one another on the support. The fusion proteins of the invention can be attached to the support by covalent or noncovalent attachment. Preferably, the support is biocompatible.

"Biocompatible," as used herein, means that the support does not generate an immune response in the subject (e.g., the production of antibodies). The support can be a biodegradable substrate carrier, such as a polymer bead or a liposome. The support can further include alum or other suitable adjuvants. The support can be a virus (e.g., adenovirus, poxvirus, alphavirus), bacteria (e.g., Salmonella) or a nucleic acid (e.g., plasmid DNA, CpG).

[00170] The dosage and frequency (single or multiple doses) administered to a subject can vary depending upon a variety of factors, including prior exposure to an antigen, a viral protein, the duration of viral infection, prior treatment of the viral infection, the route of administration of the composition or fusion protein; size, age, sex, health, body weight, body mass index, and diet of the subject; nature and extent of symptoms of viral exposure, viral infection and the particular viral responsible for the infection (e.g., influenza virus), or treatment or infection of an other antigen, such as an influenza antigen, kind of concurrent treatment, complications from the viral exposure, viral infection or exposure or other health-related problems. Other therapeutic regimens or agents can be used in conjunction with the methods and compositions and fusion proteins of the present invention. For example, the administration of the compositions and fusion proteins can be accompanied by other viral therapeutics or use of agents to treat the symptoms of a condition associated with or consequent to exposure to the antigen, such as influenza infection.

Adjustment and manipulation of established dosages (e.g., frequency and duration) are well within the ability of those skilled in the art.

[00171] Exemplary portions of antigens, flagellin and fusion proteins of the invention include those listed in Table 1.

[00172] TABLE 1

Fusion HA Strain Name/Format Insertion Site Portion of HA SEQ Protein ID

NO

HL772 B/Wisconsin/1/2010 276-277 in SEQ ID NO: 127 HA1 -2 extension, 126

D3I-01 SEQ ID NO: 50

HL657 B/Brisbane/60/2008 276-277 in SEQ ID NO: 127 HA1 -2, SEQ ID 128

D3I-01 NO: 48

HL775 A/Perth/16/2009 276-277 in SEQ ID NO: 127 HA1 -1 L modified, 129

D3I-01 SEQ ID NO: 228

Fusion HA Strain Name/Format Insertion Site Portion of HA SEQ Protein ID

NO

HL863 B/Sichuan/379/1999 276-277 in SEQ ID NO: 127 HA1 -2 extension, 182

D3I-01 SEQ ID NO: 56

HL903 B/Sichuan/379/1999 259-260 in SEQ ID NO: 2 HA1 -2 extension, 183

D3I-M SEQ ID NO: 56

HL869 B/Hubei- 276-277 in SEQ ID NO: 127 HA1-2 extension, 184

Wujiagang/158/2009 SEQ ID NO: 58 D3I-01

HL871 B/Texas/6/201 1 276-277 in SEQ ID NO: 127 HA1 -2 extension, 185

D3I-01 SEQ ID NO: 60

HL639 A/Perth/16/2009 276-277 in SEQ ID NO: 127 HA1 -1 L, SEQ ID 186

D3I-01 NO: 98

HL836 A Victoria/361/201 1 276-277 in SEQ ID NO: 127 HA1 -1 L, SEQ ID 187

D3I-01 NO: 98

HL982 A/Anhui/1/2005 276-277 in SEQ ID NO: 127 HA1 -1 L, SEQ ID 188

D3I-01 NO: 104

HL960 A/lndonesia/5/2005 276-278 in SEQ ID NO: 127 HA1 -1 L, SEQ ID 189

D3I-01 NO: 106

HL876 A/New York/107/2003 276-277 in SEQ ID NO: 127 HA1 -1 L, SEQ ID 190

D3I-01 NO: 1 12

HL926 A/Canada/Rv504/2004 276-277 in SEQ ID NO: 127 HA1 -1 L, SEQ ID 191

D3I-01 NO: 1 14

HL880 A/Netherlands/219/03 276-277 in SEQ ID NO: 127 HA1-1 L, SEQ ID 192

D3I-01 NO: 1 15

HL958 B/Hong Kong/259/2010 276-277 in SEQ ID NO: 127 HA1-2 extension, 193

D3I-01 SEQ ID NO: 54

[00173] Additional portions of antigens, flagellin and fusion proteins for use of the invention can include those listed in Table 2. For example, the fusion protein can include a portion of the globular head of HA fused to loop 6 of domain 3 of S. typhimuium FljB flagellin at the D3I-ol insertion site between residues 276-277 in SEQ ID NO: 127.

[00174] TABLE 2

HA Strain Name Format Portion of HA SEQ ID NO

A/Perth/16/2009 D3I-01 HA1 -2, SEQ ID NO: 64 194

A/Perth/16/2009 D3I-01 HA1 -1 , SEQ ID NO: 1 1 195

A/Wyoming/03/2003 D3I-01 HA1 -2, SEQ ID NO: 65 196

A/Wyoming/03/2003 STF2.D3lns.H HA1 -1 , SEQ ID NO: 12 197

A1 -1 WY03

D3I-01

A/Wyoming/03/2003 D3I-01 HA1-1 L, SEQ ID NO: 99 198

A/New York/2782/2004 STF2.D3lns.H HA1-2, SEQ ID NO: 66 199

A1 -2 NY2782

D3I-01

A/New York/2782/2004 STF2.D3lns.H HA1-1 , SEQ ID NO: 13 200 HA Strain Name Format Portion of HA SEQ ID NO

A1-1 NY2782

D3I-01

A/New York/2782/2004 D3I-01 HA1-1L, SEQ ID NO: 201

100

A/Victoria/361/2011 STF2.D3lns.H HA1-2, SEQ ID NO: 67 202

A1-2 VT361

D3I-01

A/Victoria/361/2011 STF2.D3lns.H HA1-1, SEQ ID NO: 14 203

A1-1 VT361

D3I-01

A/Aichi/2/68 STF2.D3lns.H HA1-2, SEQ ID NO: 68 204

A1-2 Aichi2

D3I-01

A/Aichi/2/68 STF2.D3lns.H HA1-1, SEQ ID NO: 15 205

A1-1 Aichi2

D3I-01

A/Aichi/2/68 D3I-01 HA1-1L, SEQ ID NO: 206

102

A/Wisconsin/67/2005 STF2.D3lns.H HA1-2, SEQ ID NO: 69 207

A1-2 WI67

D3I-01

A/Wisconsin/67/2005 STF2.D3lns.H HA1-1, SEQ ID NO: 16 208

A1-1 W!67

D3I-01

A/Wisconsin/67/2005 STF2.D3lns.H HA1-1L, SEQ ID NO: 209

A1-1L WI67 103

D3I-01

A/Anhui/1/2005 STF2.D3ins.H HA1-2, SEQ ID NO: 70 210

A1-2 AH1 D3I- 01

A/Anhui/1/2005 STF2.D3lns.H HA1-1, SEQ ID NO: 17 211

A1-1 AH1 D3I- 01

A/Bar headed STF2.D3lns.H HA1-2, SEQ ID NO: 71 212 goose/Qinghai/1 A/2005 A1-2QH1A

D3I-01

A/Bar headed STF2.D3lns.H HA1-1, SEQ ID NO: 18 213 goose/Qinghai/1 A/2005 A1-1 QH1A

D3I-01

A/Bar headed STF2.D3lns.H HA1-1L, SEQ ID NO: 214 goose/Qinghai/1 A/2005 A1-1LQH1A 105

D3I-01

A/lndonesia/5/2005 STF2.D3lns.H HA1-2, SEQ ID NO: 72 215

A1-2 IND5

D3I-01

A/lndonesia/5/2005 STF2.D3lns.H HA1-1, SEQ ID NO: 19 216

A1-1 IND5

D3I-01

A/Vietnam/1203/2004 STF2.D3lns.H HA1-2, SEQ ID NO: 80 217

A1-2 VN1203

D3I-01

A/Vietnam/1203/2004 STF2.D3!ns.H HA1-1, SEQ ID NO: 20 218

A1-1 VN1203

D3I-01 HA Strain Name Format Portion of HA SEQ ID NO

A/Vietnam/1203/2004 STF2.D3lns.H HA1-1L, SEQ ID NO: 219

A1-1L VN1203 107

D3I-01

A/Hubei/1/2010 STF2.D3lns.H HA1-2, SEQ ID NO: 73 220

A1-2 HB1 D3I- 01

A/Hubei/1/2010 STF2.D3lns.H HA1-1, SEQ ID NO: 21 221

A1-1 HB1 D3I- 01

A/Hubei/1/2010 STF2.D3lns.H HA1-1L, SEQ ID NO: 222

A1-1L HB1 108

D3I-01

A/Hong Kong/156/97 STF2.D3lns.H HA1-2, SEQ ID NO: 74 223

A1-2 HK156

D3I-01

A/Hong Kong/156/97 STF2.D3lns.H HA1-1, SEQ ID NO: 22 224

A1-1 HK156

D3I-01

A/Hong Kong/156/97 STF2.D3lns.H HA1-1L, SEQ ID NO: 225

A1-1L HK156 109

D3I-01

A/Anhui/1/2013 STF2.D3lns.H HA1-2, SEQ ID NO: 75 226

A1-2 AH1 D3I- o1

A/Anhui/1/2013 STF2.D3lns.H HA1-1, SEQ ID NO: 23 227

A1-1 AH1 D3I- 01

A/Anhui/1/2013 STF2.D3lns.H HA1-1L, SEQ ID NO: 228

A1-1L AH1 110

D3I-01

A/Turkey/ltaly/214845/2002 STF2.D3lns.H HA1-2, SEQ ID NO: 76 229

A1-2

IT214845 D3I- o1

A/Turkey/ltaly/214845/2002 STF2.D3lns.H HA1-1, SEQ ID NO: 24 230

A1-1

IT214845 D3I- 01

A/Turkey/ltaly/214845/2002 STF2.D3lns.H HA1-1L, SEQ ID NO: 231

A1-1L 111

IT214845 D3I- 01

A/New York/107/2003 STF2.D3lns.H HA1-2, SEQ ID NO: 77 232

A1-2 NY107

D3I-01

A/New York/107/2003 STF2.D3lns.H HA1-1, SEQ ID NO: 25 233

A1-1 NY107

D3I-01

A/Mallard/Netherlands/12/2 STF2.D3lns.H HA1-2, SEQ ID NO: 78 234 000 A1-2 NL12

D3I-01

A/Mallard/Netherlands/12/2 STF2.D3lns.H HA1-1, SEQ ID NO: 26 235 000 A1-1 NL12

D3I-01 HA Strain Name Format Portion of HA SEQ ID NO

A/Mallard/Netherlands/12/2 STF2.D3lns.H HA1-1L, SEQ ID NO: 236 000 A1-1L NL12 113

D3I-01

A/Canada/Rv504/2004 STF2.D3lns.H HA1-2, SEQ ID NO: 79 237

A1-2 CA504

D3I-01

ACanada/Rv504/2004 STF2.D3lns.H HA1-1, SEQ ID NO: 27 238

A1-1 CA504

D3I-01

A/Netherlands/219/03 STF2.D3lns.H HA1-2, SEQ ID NO: 81 239

A1-2 NL219

D3I-01

A/Netherlands/219/03 STF2.D3lns.H HA1-1, SEQ ID NO: 28 240

A1-1 NL219

D3I-01

A/Hong Kong/33982/2009 STF2.D3lns.H HA1-2, SEQ ID NO: 82 241

A1-2 HK33982

D3I-01

AHong Kong/33982/2009 STF2.D3lns.H HA1-1, SEQ ID NO: 29 242

A1-1 HK33982

D3I-01

A/Hong Kong/33982/2009 STF2.D3lns.H HA1-1L, SEQ ID NO: 243

A1-1L 116

HK33982 D3I- 01

A/Hong Kong/1073/99 STF2.D3lns.H HA1-2, SEQ ID NO: 83 244

A1-2 HK1073

D3I-01

A/Hong Kong/1073/99 STF2.D3lns.H HAM, SEQ ID NO: 30 245

A1-1 HK1073

D3I-01

A/Hong Kong/1073/99 STF2.D3lns.H HA1-1L, SEQ ID NO: 246

A1-1L HK1073 117

D3I-01

A/Chicken/Hong STF2.D3lns.H HA1-2, SEQ ID NO: 84 247 Kong/G9/97 A1-2 HKG9

D3I-01

A/Chicken/Hong STF2.D3lns.H HA1-1, SEQ ID NO: 31 248 Kong/G9/97 A1-1 HKG9

D3I-01

A/Chicken/Hong STF2.D3lns.H HA1-1L, SEQ ID NO: 249 Kong/G9/97 A1-1L HKG9 118

D3I-01

A/chicken/Anhui/AH 16/2008 STF2.D3lns.H HA1-2, SEQ ID NO: 85 250

A1-2 AH16

D3I-01

A/chicken/Anhui/AH 16/2008 STF2.D3lns.H HA1-1, SEQ ID NO: 32 251

A1-1 AH16

D3I-01

Achicken/Anhui/AH 16/2008 STF2.D3lns.H HA1-1L, SEQ ID NO: 252

A1-1L AH16 119

D3I-01

A/Swine/Hong Kong/9/98 STF2.D3lns.H HA1-2, SEQ ID NO: 86 253

A1-2 HK9 D3I- HA Strain Name Format Portion of HA SEQ ID NO o1

A/Swine/Hong Kong/9/98 STF2.D3lns.H HA1 -1 , SEQ ID NO: 33 254

A1 -1 HK9 D3I- o1

A/Swine/Hong Kong/9/98 STF2.D3lns.H HA1 -1 L, SEQ ID NO: 255

A1 -1 L HK9 120

D3I-01

A swine/Guangxi/58/2005 STF2.D3lns.H HA1 -2, SEQ ID NO: 87 256

A1 -2 GX58

D3I-01

A/swine/Guangxi/58/2005 STF2.D3lns.H HA1 -1 , SEQ ID NO: 34 257

A1-1 GX58

D3I-01

A/swine/Guangxi/58/2005 STF2.D3lns.H HA1 -1 L, SEQ ID NO: 258

A1 -1 L GX58 121

D3I-01

B/Shanghai/361/2002 STF2.D3lns.H HA1 -2, SEQ ID NO: 89 259

A1 -2 D3I-01

with extension

B/Malaysia/2506/2004 STF2.D3lns.H HA1 -2, SEQ ID NO: 62 260

A1 -2 D3I-01

with extension

B/Ohio/1/2005 STF2.D3lns.H HA1 -2, SEQ ID NO: 91 261

A1 -2 D3l-o1

with

extension

B/Hong Kong/330/2001 STF2.D3lns.H HA1-2, SEQ ID NO: 93 262

A1-2 D3l-o1

with extension

EXEMPLIFICATION

[00175] Example 1. Primary sequence analysis of flagellin to identify antigen insertion sites to target influenza B

[00176] Primary sequence analysis showed that the amino acid sequence of portions of influenza viral hemagglutinin, such as HA 1-2 of the Flu B (Influenza B) globular head (SEQ ID NOs: 45, 47 and 61), included a higher content of positively charged residues compared to most influenza A globular head portions, such as SEQ ID NO: 63. At neutral pH, a portion of an influenza Flu B globular head, such as SEQ ID NO: 45, has a net positive charge with a pi of between about 8.5 to about 9.5. At neutral pH, flagellin has a net negative charge due, in part, to a series of negatively charged amino acids in Domain 1(D1) of flagellin with a pi of about 4.5 to about 5.0. Apposition of a positively charged portion of an influenza B HA with the negatively charged Dl of flagellin, such as in R3 and R3.2x fusion protein, may result in intra-molecular interactions between the portion of HA and portion of flagellin that may interfere with the ability of flagellin to trigger TLR5 signaling and with the presentation of the influenza B HA antigen to immune cell receptors.

[00177] Immunogenicity of an R3 fusion protein of composition that includes an influenza B, such as a B/Florida/4/2006 as in HL098 (SEQ ID NO: 122), may be improved by extending the portion of the globular head amino acid sequence by an additional 9 amino acids (SEQ ID NO: 123) at the C-terminus of HA 1-2. For example, an additional 9 amino acids of the HA (SLPLIGEAD (SEQ ID NO: 123), which corresponds to native amino acid residues 300-308 of SEQ ID NO: 124, can be added to the carboxy-terminus of HA1-2 (e.g., SEQ ID NO: 45) to generate SEQ ID NO: 46. For example, the HA1-2 with an extension can then be used to generate a fusion protein, such as HL352 (SEQ ID NO: 125). The additional 9 amino acids to the portion of HA (e.g., SEQ ID NO: 123) introduced two negatively charged amino acids (E306 and D308) to the portion of the globular head, which may serve to "repel" the portion of the globular head away from the negatively charged flagellin. However, as described infra, while improved, the activity (TLR5 activation and immunogenicity) associated with fusion proteins having HA components with the additional 9 amino acids was low, as discussed infra with respect to, for example, FIGs. 3 and 5 and fusion proteins referred to as "HL098" (SEQ ID NO: 122) and "HL352" (SEQ ID NO: 125).

[00178] Fusion of an antigen with an isoelectric point greater than about 7.0 to a loop of Domain 3 of flagellin is believed to position the HA antigen component of the fusion protein distant (i.e., away) from domain (Dl) (the TLR5 -binding domain) of flagellin. Portions of an influenza viral hemagglutinin that include at least a portion of a globular head having at least one β-sheet at the bottom of the globular head and that have an isoelectric point greater than about 7.0, such as an influenza B viral hemagglutinin (e.g., SEQ ID NO: 45, 61), can be fused within at least one loop of domain 3 of flagellin to form a "D3Ins format," also referred to as a "D3I" or "D3I construct." Likewise, portions of an influenza viral hemagglutinin that include at least a portion of a globular head having at least one β-sheet at the bottom of the globular head and that have an isoelectric point greater than about 7.0, such as an influenza A viral hemagglutinin of SEQ ID NO: 98, and A/Perth/16/2009 strains, can be fused within at least one loop of domain 3 of flagellin to form a "D3Ins format." Fusion proteins that include both influenza A and influenza B antigens can be combined to form multivalent compositions that target both influenza A and influenza B antigens to thereby prevent or minimize disease associated with influenza A and influenza B infection.

[00179] The crystal structure of domain 3 of Salmonella typhimurium flagellin (FliC, Protein Data Bank ID (PDB): lUCU) has been reported (Yonekura, K., et al, Nature, 424: 643-650 (2003)) and includes the identification of 6 loops in domain 3 of SEQ ID NO: 1. Secondary structure prediction of Salmonella typhimurium flagellin FliC using PHD is depicted below and, consistent with the known tertiary structure, predicts 6 loops in domain 3. Designations below are AA: Primary amino acid sequence; P OF_sec: Secondary structure prediction where "FT stands for ot- Helix and "E" stands for β-strand (Rost, B., et al, Proteins, 19: 55-72 (1994)). The predicted loops in domain 3 of SEQ ID NO: 1 are indicated in boxes below. The secondary structures adjacent to the predicted loops are essentially similar to the known secondary structures for Salmonella typhimurium flagellin (FliC, Protein Data Bank ID (PDB): lUCU) (Yonekura, K., et al, Nature, 424: 643-650 (2003)).

[00180] The predicted secondary structure of FliC (5". typhimurium, SEQ ID NO: 1) is substantially similar to the known high resolution atomic model determined by combination of X-ray crystallography and electron microscopy (lUCU, Yonekura, K., et al. Nature, 424: 643-650 (2003)). The substantial similarity between the predicted secondary structure and the known secondary structure of S. typhimurium FliC (SEQ ID NO: 1) indicates that secondary structures predicted employing CLUSTALW and PHD adjacent to loops of domain 3 of other flagellins, including S. typhimurium FljB (SEQ ID NO: 2), can be employed to select insertion sites that correspond to known loops in domain 3 in other flagellin, such as of S. typhimurium FliC.

[00181] Secondary structure prediction using PHD of Flagellin FliC (Salmonella typhimurium, Protein Data Bank ID: lUCU, SEQ ID NO: 1) is depicted below. Predicted loops in domain 3 are indicated by boxed text. ....,....1....,....2....,....3....,....4....,....5....,....6

AA AQVINTNSLSLLTQNNLNKSQSALGTAIERLSSGLRINSAKDDAAGQAIA RFTANI GL PROF sec EEE HHHHHHHHHHHHHHHHHHHHHHHH EE HHHHHHHHHHHHHHHHH

....,....7....,....8....,....9....,....10...,....11...,....12

AA TQASRNANDGISIAQTTEGALNEINNNLQRVRELAVQSANSTNSQSDLDSIQAEITQRLN PROF sec HHHHHHHH HHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHHHHHHHHHHH

.........13........14........15........16........17........18

AA EIDRVSGQTQFNGVKVLAQDNTLTIQVGANDGETIDIDLKQINSQTLGLDTLNVQQKYKV PROF_sec HHHHHHH EEEEE EEEEEE EEEEEEEEEE EEEEEEEEEEE

.........19........20........21... , ...22. . , ....23... , ..■ .24

AA SDTAATVTGYADTTIALDNSTFKASATGLG(GT|DQKI|DGD|LKF1DDTTGKY]YAKVTVT|GGTG|

PROF sec EEEE EEEEE EEE EEE EEEEEEEEEEE

.........25... , ....26........27 ....28........29........30

AA [KD|GYYEVSV|DKTNGE^"|VTL[AGGATSPLTGGLPATAT|EDVKNVQVANADLTEAKAALTAAGV

PROF sec EEEEEEE ΕΕΕΕΕΕΕΕ EEEE EEE

....,....31...,....32...,....33...,....34...,....35...,....36

AA TGTASWKMSYTDNNGKTIDGGLAVKVGDDYYSATQ KDGSISINTTKYTADDGTSKTAL

PROF sec EEEE EEEE EEE EEEE EE EEE

....,....37...,....38...,....39...,....40...,....41...,....42

AA NKLGGADGKTEWSIGGKTYAASKAEGHNFKAQPDLAEAAATTTENPLQKIDAALAQVDT

PROF sec EE EEEE EEE HHHHHHHHHHHHHHHHHHHH

....,....43...,....44...,....45...,....46...,....47...,....48

AA LRSDLGAVQNRFNSAITNLGNTVNNLTSVRSRIEDSDYATEVSNMSRAQILQQAGTSVLA PROF sec HHHHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHHHHHHHHHHHHHHHHHHHH

, 49...

AA QANQVPQNVLSLLR (SEQ ID NO: 1)

PROF_sec HH HHHHHHH (predicted secondary structure)

[00182] Secondary structure prediction using PHD of Flagellin FljB (S.

typhimurium, SEQ ID NO: 2). Predicted loops in domain 3 are indicated by boxed text.

....,....1....,....2....,....3....,....4....,....5....,....6

AA MAQVINTNSLSLLTQNNLNKSQSALGTAIERLSSGLRINSA DDAAGQAIANRFTANIKG PROF sec EEEE HHHHHHHHHHHHHHHHHHHHHHHH EE HHHHHHHHHHHHHHHH

....,....7....,....8....,....9....,....10...,....11...,....12

AA LTQASRNANDGISIAQTTEGALNEINNNLQRVRELAVQSANSTNSQSDLDSIQAEITQRL PROF sec HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHHHHHHHHHH

.........13........14........15........16........17........18

NEIDRVSGQTQFNGVKVLAQDNTLTIQVGANDGETIDIDLKQINSQTLGLDSLNVQKAYD HHHHHHHH EEEE EEEEE EEEEEEEEE EEEEEEEEEEEEE

.........19........20........21........22... , ...23.■■ , ....24

VKDTAVTTKAYANNGTTLDVSGLDDAAIKAATGGTN|G^SVT|GGA|VKF1DADNNKY|FVTIG EEE EEEEEEE EEEEE EEE EEEEE EEEEEEE .... , ....25........26........27........28........29... , ....30

AA QFT|GADAAl^GDYEAnSfV|¾TDGT|VTL|AAGATKTTMPAGATTKT|BVQELKDTPAVVSADAiaSf PROF sec E EEEEE EEEE EEE

.... , ....31... , ....32... , ....33... , ....34... , ....35... , ....36

AA ALIAGGVDATDANGAELVK SYTDKNGKTIEGGYALKAGDKYYAADYDEATGAIKAKTTS PROF sec EEEE EEEEEE EE EEE EEE

.........37........38... , ....39........40........41... , ....42

AA YTAADGTTKTAANQLGGVDGKTEWTIDGKTYNASKAAGHDPKAQPELAEAAAKTTENPL PROF sec EEEEEEEEE EEEE EE HHHHHHHH

.... , ....43... , ....44... , ....45... , ....46... , ....47... , ....48

AA QKIDAALAQVDALRSDLGAVQNRFNSAITNLGNTVWNLSEARSRIEDSDYATEVSNMSRA PROF_sec HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHH HHHHHHHHHHHHH

.........49........50... , .

AA QILQQAGTSVLAQA QVPQNVLSLLR (SEQ ID NO : 2)

PROF_sec HHHHHHHHHHHHHH HHHHHHH (predicted secondary structure)

[00183] The predictions of loops in domain 3 employing PHD compared to known loops identified by X-ray crystallography show that the PHD computer program can be employed to predict amino acid sequences and loops for sites of insertion of antigens within domain 3 of a flagellin.

[00184] Sequence alignment between S. typhimurium FliC (SEQ ID NO: 1) and S. typhimurium FljB (SEQ ID NO: 2) was performed using the multiple sequence alignment tool CLUSTALW and secondary structure prediction by PHD with PROF_sec: structure prediction where "H" stands for alpha-helix and "E" stands for beta-strand (Thompson, J.D., et al., Nucleic Acids Res.22: 4673-4680 (1994); Rost, B., et al., Proteins 19: 55-72 (1994)). The sequence alignment below showed about 74.75% identity of amino acid residues denoted with (*) and about 10.26% difference with about 7.30%> strongly similar (:) and about 7.69% (.) weakly similar amino acid residues. Domain boundaries of DO, DJL, D2 and Ώ1 are underlined differently and three D3 insertion sites were marked.

[00185] The objective of this alignment was to predict loop regions in Domain 3 of S. typhimurium FljB (SEQ ID NO: 2) that correspond to known loop regions of Domain 3 of S. typhimurium FliC (SEQ ID NO: 1) for points of insertion of portions of HA that have an isoelectric point greater than about 7.0 to generate fusion proteins of the inventions. [00186] Primary amino acid sequence alignment between FliC (PDB: lUCU (Yonekura, K., et al, Nature, 424: 643-650 (2003)), SEQ ID NO: 1) and FljB (SEQ ID NO: 2) of Salmonella typhimurium indicated that the two flagellins shared highly conserved DO (domain 0) and Dl (domain 1) domains, but varied in the D2 (domain 2) and D3 (domain 3) domains. However, secondary structure prediction, using PHD (Rost, B., et al, Proteins 19: 55-72 (1994)), showed that both D2 and D3 in the FliC flagellin FliC (SEQ ID NO: 1) and FljB flagellin (SEQ ID NO: 2) of

Salmonella typhimurium share similar secondary structures, despite the differences in primary amino acid sequence.

[00187] Sequence alignment and secondary structure prediction using

CLUSTALW and PHD of FliC (SEQ ID NO: 1) and FljB (SEQ ID NO: 2) of S. typhimurium is depicted below. Predicted loops in domain 3 are indicated by boxed text.

10 20 30 40 50 60

FliCxx0 MAQVINTNSLSLLTQNNLNKSQSALGTAIERLSSGLRINSAKDDAAGQAIANRFTANIKG lUCU HHHHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHHHHHHHHHHH

PROF_sec EEE HHHHHHHHHHHHHHHHHHHHHHHH EE HHHHHHHHHHHHHHHH

FlijBxl MAQVINTNSLSLLTQNNLNKSQSALGTAIERLSSGLRINSAKDDAAGOJ^IA RJ ANIKG

PROF_sec EEEE HHHHHHHHHHHHHHHHHHHHHHHH EE HHHHHHHHHHHHHHHH

****************************************

70 30 30 100 110 120

FliCxxO LTQASRNANDGISIAQTTEGALNEIN NLQRVRELAVQSANSTNSQSDLDSIQAEITQRL lUCU HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHHHHHHHHHHH PROF_sec HHHHHHHH HHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHHHHHHHHHH FlijBxl

PROF sec HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHHHHHHHHHH

************************************************************

130 140 150 160 170 180

FliCxxO NEIDRVSGQTQFNGVKVLAQDNTLTIQVGA DGETIDIDLKQINSQTLGLDTLNVQQKYK lUCU HHHHHHHHHH EEEEEE EEEEEE E PROF_sec HHHHHHHH EEEEE EEEEEE EEEEEEEEEE ΕΕΕΕΕΕΕΞΕΕΕ FlijBxl

PROF sec HHHHHHHH EEEE EEEEE EEEEEEEEE EEEEEEEEEEEEE

*************************************************** . **** . *

190 200 210 220 230 240

I I I I I I

vsDTAAT GYADT TIALDNSTFKASATGLG|G^DQKI|DGD|LKF|DDTTGKY|YAKVT

EEEEE EEEE EEEE HHHH HHHHH EEEE EEE EEEEEE

EEEE EEEEE EEE EEE EEEEEE VKpTAyTTKAYANNGTT^ 250 260 270 280 290 300

FliCxxO -ντ|θΟ-Τΰκ ΥΥΕν3ν[ΡΚΤΝΟΕ|νΤί|ΑθαΑΤ3ΡίΤΟΟίΡΑΤΑΤ|ΕΡν ΝνθνΑΝΑΡίΤΕΑΚ lUCU EEEEEE EEE EEE EEEEE HHHHH

PROF_sec EEEE E EEEEEEE EEEEEEEE EEEE

PlijBxl GFT|GADAAK^GDYETOV]A-TDGT|VTL|AAGATKTTMPAGATT^EVQELKDTPAVVSADAK

PROF sec E EEEEE EEEE EEE

*.* *** * *.* **** ***

D3I site t T

310 320 330 340 350 360

FliCxxO AALTAAGVTGTAS WKMSYTDNNGKTIDGGLAVKVGDDYYSATQNKD-GSISINTT lUCU HHHHHH EE EEEEEEE EEEEEEEEEE EEEE EEE PROF_sec EEE E EEE EEEE EEE EEEE FlijBxl NALIAGGVDATDANGAELV MSYTDKNGKTIEGGYALKAGDKYYAADYDEATGAIKAKTT PROF sec EEEE EEEEEE EE EEE EEE

* * * * * .*******.*****.** *.* ** **.*

370 380 390 400 410 420

FliCxxO KYTADDGTS TALNKLGGADGKTEWSIGGKTYAASKAEGHNFKAQPDLAEAAATTTENP lUCU EEE EEEEEE EEEE HHHH HH PROF_sec EE EEEEE EEEE EEE HHHHHHH FlijBxl SYTAADGTTKTAANQLGGVDGKTEWTIDGKTYNASKAAGHDFKAQPELAEAAAKTTENP PROF sec EEEE EE HHHHHHH

*** ***.*** *.*** *******.* **** **** **.*****.****** *****

430 440 450 460 470 480

FliCxxO LQKIDAALAQVDTLRSDLGAVQNRFNSAITNLGNTVNNLTSVRSRIEDSDYATEVSNMSR lUCU HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHHHH PROF_sec HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHHHHHHH FlijBxl

PROF sec HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHH HHHHHHHHHHHH

************ . ************************** . ******************

490 500

FliCxxO AQILQQAGTSVLAQANQVPQ VLSLLR (SEQ ID NO : 1)

lUCU HHHHHHHHHHHHHHHHHHHHHHHH (empirical structure)

PROF_sec HHHHHHHHHHHHHHH HHHHHHH (predicted secondary structure) FlijBxl AQILQQAGTSVLAQANQVPQNVLSLLR (SEQ ID NO: 2)

PROF sec HHHHHHHHHHHHHHH HHHHHHH (predicted secondary structure)

***************************

[00188] An insertion site in a predicted loop in domain 3 of S. typhimurium FljB, referred to herein as "D3I-ol," for fusion of an antigen that has an isoelectric point greater than about 7.0, is between amino acid residues G277 and A278 of SEQ ID NO: 2. The insertion site D3I-ol is between predicted β-strands VTLA (SEQ ID NO: 131), which is amino acid residues 263-266 of SEQ ID NO: 2, and VVS, which is amino acid residues 293-295 of SEQ ID NO: 2. Another insertion site in a second predicted loop in domain 3 of S. typhimurium FljB, referred to herein as "D3I-il," is between residues T259 and D260 of SEQ ID NO: 2. The insertion site D3I-il is between β-strands EVNVA (SEQ ID NO: 132), which corresponds to amino acid residues 254-258 of SEQ ID NO: 2, and VTLA (SEQ ID NO: 131), which corresponds to amino acid residues 263-266 of SEQ ID NO: 2. An additional insertion site in a third predicted loop in domain 3 of S. typhimurium FljB, referred to herein as "D3I-sl," is between residue G244 and A245 of SEQ ID NO: 2. The insertion site of D3I-sl is between β-strands KYFVTIGG (SEQ ID NO: 133), which corresponds to amino acid residues 234-241 of SEQ ID NO: 2, and EVNVA (SEQ ID NO: 132), which corresponds to amino acid residues 254-258 of SEQ ID NO: 2. The secondary structures adjacent to the predicted loops of S. typhimurium FliC are substantially similar to the predicted secondary structures adjacent to loops of domain 3 of S. typhimurium FljB and are depicted in FIG. 29.

[00189] In an embodiment, fusion of the antigen with an isoelectric point greater than about 7.0 can be about 2 to about 10 amino acids towards the carboxy- or amino-terminus of flagellin from the designated insertion site, based on the proximity of the adjacent secondary structural elements. For example, the D3I-ol site can be from amino acid residues G268 through D289 of SEQ ID NO: 2, which does not invade adjacent β-strands, and are predicted at VTLA (SEQ ID NO: 131) at amino acid residues 263-266 of SEQ ID NO: 2 and VVS at amino acid residues 293-295 of SEQ ID NO: 2. In addition, for example, the D3I-il site can only accommodate a shift of 1 or 2 amino acids towards the carboxy-terminus of flagellin of SEQ ID NO: 2 at amino acid residues D260 or G261 of SEQ ID NO: 2 before disrupting the neighboring β-strands, which are predicted at EVNVA (SEQ ID NO: 132) at amino acid residues 254-258 of SEQ ID NO: 2) and VTLA (SEQ ID NO: 131) at amino acid residues 263-266 of SEQ ID NO: 2.

[00190] Based on predicted secondary structure analysis, loops of domain 3 of E. coli FliC flagellin (SEQ ID NO: 3) and Vseudomonas aeruginosa P^O flagellin (SEQ ID NO: 4) can be predicted. Based on this analysis, E. coli FliC primary amino acid sequence (SEQ ID NO: 3) was aligned with the primary amino acid sequence of Samonella typhimurium FliC (SEQ ID NO: 1). The secondary structures were assigned either directly from known structure of lUCU for Samonella typhimurium FliC, (SEQ ID NO: 1) or by prediction using PHD. The secondary structures of E. coli flagellin (SEQ ID NO: 3) or Pseudomonas aeruginosa PAOl flagellin (SEQ ID NO: 4) were assigned using the PHD program. Designations depicted below are AA: Primary amino acid sequence; PROF_sec: Secondary structure prediction where "H" stands for a-Helix and "E" stands for beta-strand (Rost, B., et al, Proteins 19: 55-72 (1994)).

[00191] Secondary structure prediction using PHD of Flagellin FliC (E. coli, SEQ ID NO: 3) is depicted below. Predicted loops in domain 3 are in boxed text.

.........1.........2.........3.........4.........5.........6

AA MAQVINTNSLSLLTQN LNKSQSSLSSAIERLSSGLRINSAKDDAAGQAIANRFTANIKG

PROF sec EEEE HHHHHHHHHHHHHHHHHHHHHHHH EE HHHHHHHHHHHHHHHH

.........7.........8.........9.........10........11........12

AA LTQASRNANDGISVAQTTEGALNEINNNLQRIRELSVQATNGTNSDSDLSSIQAEITQRL

PROF sec HHHHHHHHH HHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHHHHHHHHHH

.........13........14........15........16........17........18

AA EEIDRVSEQTQFNGVKVLAENNEMKIQVGANDGETITINLAKIDAKTLGLDGFNIDGAQK

PROF sec HHHHHHHH EEEEE EEEEE · EEEEEEEEE EEEEEEEEEE

.........19........20........21........22. .23........24

AA ATGSDLI sKFKATGTDNYDVGGKTYTVWESGAVP^DA K|DV|FVSA[AD|GSLTTSSDTKV|S]

PROF sec EEEEEE EEEEE ΕΕΕΕΕΕΕ EEEEEE EEEEEE EE

.........25...,....26........27........28........29..., ....30

AA |GES|IDATELAKLAIKL|ADKGS|IEYKGITF[TNNTGAELDA G|KGVLTA IDGQDVQFTIDS

PROF sec EEE EEEEEEEEE ΕΕΕΕ EEE EEEEEE EEEEEE

.........31........32........33........34........35........36

AA APTGAGATITTDTAVYKNSAGQFTTT VENKAATLSDLDLNAAKKTGSTLWNGATYNV

PROF sec EEE EEEEE EEEEE EEE EEEEEEEEEEEE

.........37........38........39........40...,....41........42

AA SADGKTVTDTTPGAPKVMYLSKSEGGSPILV EDAAKSLQSTTNPLETIDKALAKVDNLR

PROF sec E EEE EEEEE EEE HHHHHHHHHHHHHHHHHHHHH

.........43...,....44........45........46........47........48

AA SDLGAVQNRFDSAITNLGNTVNNLSSARSRIEDADYATEVSNMSRAQILQQAGTSVLAQA

PROF sec HHHHHHHHHHHHHHHHHHHHHH HHHHHHH HHHHHHHHHHHHHHHHHHHHHHHHHHH

.........49.

AA NQTTQNVLSLLR (SEQ ID NO : 3)

PROF sec HHHHHHH (predicted secondary structure)

[00192] Secondary structure prediction using PHD of flagellin (P. aeruginosa PAOl, SEQ ID NO: 4) is depicted below. Predicted loops in domain 3 are indicated by boxed text. .........1.........2.........3.........4.........5. . 6

AA MALTVNTNIASLNTQRNLNASSNDLNTSLQRLTTGYRINSAKDDAAGLQISNRLSNQISG

PROF sec EEEE HHHHHHHHHHHHHHHHHHHHHHHH EE HHHHHHHHHHHHHHHH

.........7.........8.........9.........10........11........12

AA LNVATRNA DGISLAQTAEGALQQSTNILQRIRDLALQSA GSNSDADRAALQKEVAAQQ

PROF sec HHHHHHHHHH HHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHHHHHHHHHH

.........13........14........15........16........17........18

AA AELTRISDTTTFGGRKLLDGSFGTTSFQVGSNAYETIDISLQNASASAIGSYQVGSNGAG

PROF sec HHHHHHHH EEEEE EEEE EEEEEEEEE EEE EEEEE

.........19........20........21........22........23..., ....24

AA TVASVAGTATASGIASGTVNLVGGGQV NIAI AGDSA|KA|IAEKM|DGA|IPNL[SARAR jvF

PROF sec EEEEE EEEEEEEEEE EEEEE EEEEEE EEE

.... , ....25........26... , ....27........28........29........30

AA TADVS|GVTGGS|LNFDVTVjGSNT|VSL|AGVTSTQDLADQ|LNSNSSKLGI ASIND GVLTIT

PROF sec E EEE EEEEEEEEE EEEEE EEEEEEE EEEEE

.........31........32........33........34........35...,....36

AA SATGE VKFGAQTGTATAGQVAVKVQGSDGKFEAAAKNWAAGTAATTTIVTGYVQLNSP

PROF sec EE EE EEEEEEEE EEEE EEEE EEEEEEEEEE

..........37........38........39........40........41...,....42

AA TAYSVSGTGTQASQVFGNASAAQKSSVASVDISTADGAQNAIAWDNALAAIDAQRADLG

PROF sec EEEEE EEEE EE EE HHHHHHHHHHHHHHHHHHHHHHHHHH

.........43..., ....44. .45........46........47..., ....4£

AA AVQNRFK TIDNLTNISENATNARSRIKDTDFAAETAALSK QVLQQAGTAILAQANQLP PROF sec HHHHHHHHHHHHHHHHHH HHHHHHH HHHHHHHHHHHHHHHHHHHHHHHHHHH

AA QAVLSLLR (SEQ ID NO: 4)

PROF sec HHHHHHH (predicted secondary structure)

[00193] The E.coli FliC primary amino acid sequence (SEQ ID NO: 3) was initially aligned with the primary amino acid sequence of Salmonella typhimurium FliC (SEQ ID NO: 1). Sequence alignment was performed employing the multiple sequence alignment tool CLUSTALW and secondary structure prediction by PHD with PROF_sec: denoting secondary structure prediction where "H" stands for alpha-helix and "E" stands for β-strand (Thompson, J.D., et al., Nucleic Acids Res, 22: 4673-4680 (1994); Rost, B., et al, Proteins 19: 55-72 (1994)). The sequence alignment showed about 53.98% identity of amino acid residues denoted with (*) and about 21.91% difference with about 12.95% strongly similar (:) and about 11.16%) (.) weakly similar amino acid residues. The predicted secondary structure of E. coli FliC (ECFlic, SEQ ID NO: 3) was substantially similar to that of S.

typhimurium FliC (STFliC, Protein Data Bank ID: lUCU (Yonekura, K., et al, Nature 424:643-650 (2003)) , SEQ ID NO: 1) despite differences in primary amino acid sequences.

Sequence alignment and secondary structure prediction using CLUSTALW and PHD of S. typhimurium FliC (SEQ ID NO: 1) and E. coli FliC (SEQ ID NO: 3) to select loops in domain 3 for fusion for to antigens is depicted below. The predicted loops in domain 3 are indicated by boxed text.

10 20 30 40 50 60

STFliC MAQVINTNSLSLLTQNNLNKSQSALGTAIERLSSGLRINSAKDDAAGQAIANRFTANIKG lUCU HHHHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHHHHHHHHHHH

PROF_sec EEE HHHHHHHHHHHHHHHHHHHHHHHH EE HHHHHHHHHHHHHHHH

ECFlic AQVINTNSLSLLTQNNLNKSQSSLSSAIERLSSGLRINSAKDDAAGQAIANRFTA IKG PROF_sec EEEE HHHHHHHHHHHHHHHHHHHHHHHH EE HHHHHHHHHHHHHHHH

*********************** . * . *********************************

70 80 90 100 110 120

STFliC LTQASRNA DGISIAQTTEGALNEINNNLQRVRELAVQSANSTNSQSDLDSIQAEITQRL lUCU HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHHHHHHHHHHH

PROF_sec HHHHHHHH HHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHHHHHHHHHH

ECFlic LTQASRNA DGISVAQTTEGALNEINNNLQRIRELSVQATNGTNSDSDLSSIQAEITQRL PROF_sec HHHHHHHHH HHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHHHHHHHHHH

*************.*****************.***.**..* ***.*** **********

130 140 150 160 170 180

NEIDRVSGQTQFNGVKVLAQDNTLTIQVGANDGETIDIDLKQINSQTLGLDTLNVQ- -QK HHHHHHHHHH EEEEEE EEEEEE

HHHHHHHH EEEEE EEEEEE EEEEEEEEEE EEEEEEEEEEE

EEIDRVSEQTQFNGVKVLAENNEMKIQVGANDGETITINLAKIDAKTLGLDGFNIDGAQK EEIDRVSEQTQFNGVKVLAENNEMKIQVGANDGETITINLAKIDAKTLGLDGFNIDGAQK HHHHHHHH EEEEE EEEEE EEEEEEEEE EEEEEEEEEE

190 200 210 220 230 240

YKVSDTAATVTGYADTTIALDNSTFKASAT-GLG|GT|DQKI|DGD|LKF|DDTTGKY|YAKVTVT EEEEEE EEEEEEEE HHHH HH HHH EEEE EEE EEEEEE

EEEE EEEEE EEE EEE EEEEEEEE

ATGSDLISKFKATGTDNYDVGGKTYTWVESGAA^DANK|D^FVSA[ D|GSLTTSSDTKV|S| EEEEEE EEEEE ΕΕΕΕΕΕΕ EEEEEE EEEEEE EE

250 260 270 280 290 300

STFiic |G-GTGKD|GYYEVsv|DkTN-G|EVTL[AGGATSPLTGGLPATAT|EDVKNVQVANADLTEAKAA lUCU EEEEEE EEE EEE EEEEE HHHHHHH

PROF_sec EE E EEEEEEE EEEEEEEE EEEEEE

ECFlic |GES|IDATELAKLAIKL|ADKGS|IEYKGITF|T TG—AELDA G|KGVLTANIDGQDVQFT

PROF sec EEE EEEEEEEEE EEEE EEE EEEEEE ΕΕΕΕΕ D3I site

310 320 330 340 350 360

STFlic LTAAGVTGTASWKMSYTDNNGKTIDGGLAVKVGDDYYSATQNKDGSISINTTKYTADDG lUCU HHHH EEEEEEEEE EEEEEEEEEE EEEE EEE

PROF_sec E EEEE ΞΕΞΕ EEE EEEE EE ECFlic IDSNAPTGAGATIT TDTAVYKNSAGQFTTTKVENKAATLS DLDLNAAK TGSTL PROF sec E EEE EEEEE EEEEE EEE EEE

370 380 390 400 410 420

STFlic TS TALNKLGGADG TEWSIGG- -KTYAASKAEGHNFKAQPDLAEAAATTTENPLQKID lUCU EEE EEEEEE EEEE HHHH HH HHH PROF_se EEEEE EEEE EEE HHHHHHH HHH ECFlic VWGATYNVS-ADGKTVTDTTPGAPKVMYLSKSEGGSPILVWEDAAKSLQSTTNPLETID

WNGATYNVS ADGKTVTDTTPGAPKVMYLSKSEGGSPILVNEDAAKSLQSTTNPLETID

PROF se EEEEEEEEEE EEE EEEEE EEE HHHHHHHHHHH

430 440 450 460 470 480

STFlic AALAQVDTLRSDLGAVQNRFNSAITNLGNTVNNLTSARSRIEDSDYATEVSNMSRAQILQ lUCU HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHHHHHHH PROF_sec HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHHHHHHHHHH ECFlic KALAKVDNLRSDLGAVQNRFDSAITNLGNTVNNLSSARSRIEDADYATEVSNMSRAQILQ PROF sec HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHH HHHHHHHHHHHHHHHHH

*** . ** ************ . ************* . ******** . ****************

490 500

STFlic QAGTSVLAQANQVPQNVLSLLR (SEQ ID NO: 1)

lUCU HHHHHHHHHHHHHHHHHHH (empirical structure)

PROF_sec HHHHHHHHHH HHHHHHH (predicted secondary structure)

ECFlic QAGTSVLAQANQTTQNVLSLLR (SEQ ID NO: 3)

PROF sec HHHHHHHHHH HHHHHHH (predicted secondary structure)

************ ********

[00195] Sequence alignment of Salmonella typhimurium FliC (STFlic, Protein Data Bank ID: lUCU, SEQ ID NO: 1) and Pseudomonas aeruginosa PAOl flagellin (SaFlixl, SEQ ID NO: 4) was performed by using the multiple sequence alignment tool CLUSTALW and secondary structure prediction by PHD with PROF_sec: denoting secondary structure prediction where "H" stands for alpha-helix and "E" stands for beta-strand (Thompson, J.D., et al., Nucleic Acids Res. 22: 4673-4680 (1994); Rost, B., et al, Proteins 19: 55-72 (1994)). The sequence alignment showed about 36.24% identity of amino acid residues denoted with (*) and about 30.10% difference with about 20.59% strongly similar (:) and about 13.07% (.) weakly similar amino acid residues. [00196] Sequence alignment and secondary structure prediction using

CLUSTALW and PHD of S. typhimurium FliC (SEQ ID NO: 1) and P. aeruginosa PAOl flagellin (SEQ ID NO: 4) to select loops in domain 3 of the flagellin for fusion to antigens is depicted below. The predicted loops in domain 3 are indicated by boxed text.

10 20 30 40 50 60

STPlic MAQVINTNSLSLLTQN LNKSQSALGTAIERLSSGLRINSAKDDAAGQAIA RFTANIKG lUCU HHHHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHHHHHHHHHHH

PROF_sec EEE HHHHHHHHHHHHHHHHHHHHHHHH EE HHHHHHHHHHHHHHHH

SaFlixl MALTTOTNIASLNTQRNLNASSNDLNTSLQRLTTGYRINSAKDDAAGLQISNRLSNQISG

PROF sec EEEE HHHHHHHHHHHHHHHHHHHHHHHH EE HHHHHHHHHHHHHHHH

* * * * * * * • * * .. * ***********

70 90 100 110 120

LTQASRNANDGISIAQTTEGALNEIN NIiQRVRELAVQSANSTNSQSDLDSIQAEITQRL HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHHHHHHHHHHH HHHHHHHH HHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHHHHHHHHHH LNVATRNANDGISLAQTAEGALQQSTNILQRIRDLALQSA GSNSDADRAALQKEVAAQQ HHHHHHHHHH HHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHHHHHHHHHH

130 140 150 160 170 180

NEIDRVSGQTQFNGVKVLAQDN-TLTIQVGA DGETIDIDLKQINSQTLGLDTLNVQQKY HHHHHHHHHH E EEEEE EEEEEE

HHHHHHHH EEEEE EEEEEE ΕΕΞΕΕΕΕΕΕΕ ΕΕΕΕΕΕΞΕΕΕΕ

AELTRISDTTTFGGRKLLDGSFGTTSFQVGSNAYETIDISLQNASASAIGSYQVGSNGAG HHHHHHHH EEEEE EEEE EEEEEEEEE EEE EEEEE

190 200 210 220 230 240

KVSDTA- -ATVTGYADTTIALDN STFKASATGLG|GT|DQKI|DGD1LKF|DDTTGKY|YA

EEEEEE EEEEEEEE H HHH HHHHH EEEE EEE EEE

EEEE EEEEE EEE EEE EEE

TVASVAGTATASGIASGTVNLVGGGQVKNIAIAAGDSA|^IAEKM|DG^IPNL|SARART|VF EEEEE ΕΕΕΕΕΕΕΕΕΕ EEEEE EEEEEE EEE

,*: , ,* **. :* *. *: * , . *: . . . .*.** . _{< ;}

2V222AGTAT22G2A22T22L22GGQV 2222A2222222222K2DG2222222222222

250 260 270 280 290 300

STFlic KVTVT|GGTGKD|GYYEVS^DKTNGE|VTL|AGGATSPLTGGLPATAT|EDVKNVQVANADLTEA lUCU EEE EEEEEE EEE EEE EEEEE HHHH

PROF_sec EEEEEEEE EEEEEEE EEEEEEEE EEEE

SaFlixl TADVS|GVTGGS|LNFDVTV|GSN- -ijvsL|AG VTSTQDLADQ|LNSNSSKLGI ASIND

PROF_sec E EEE EEEEEEEEE EEEEE EEEEEEE

*.* ** · .*.* *.*** .* . . .

D3I site t †

310 320 330 340 350 360

STFlic KAALTAAGVTGTASWKMSYTDNNGKTIDGGLAVKVG- -DDYYSATQN DGSISINTTKY lUCU HHHHHHH EEEEEEEEE EEEEEEEEEE EEEE EEE

PROF sec EEE EEEE EEEE EEE EEEE E SaPlixl KGVLTITSATGEN VKFGAQTGTATAGQVAVKVQGSDGKFEAAAKNVVAAGTAATTT PROF sec EEEEEEE EE EEEEEEEE EEEE EEEE E

* ** . * * . . . * . * .**** * . . *

370 380 390 100 410 420

STFlic TADDGTSKTALNKLGGADGKTEWSIGGKTYAASKAEGHNFKAQPDLAEAAATTTENPLQ

FliCxxO TADDGTS TALNKLGGADGKTEWSIGGKTYAAS AEGHNFKAQPDLAEAAATTTENPLQ lUCU EEE EEEEEE EEEE HHHH HHHH

PROF_sec E EEEEE EEEE EEE HHHHHHHHH SaFlixl IVTGYVQLNSPTAYSVSGTGTQASQVFGNASAAQKSS VASVDISTADGAQNAIA PROF sec EEEEEEEEE EEEEE EEEE EE EE HHHHHHHHH

*. . *.. ** *. . .. * .

430 440 450 460 470 480

STFlic KIDAALAQVDTLRSDLGAVQNRFNSAITNLGNTVNNLTSARSRIEDSDYATEVSNMSRAQ lUCU HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHHHHHH PROF_sec HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHHHHHHHHH SaFlixl WDNALAAIDAQRADLGAVQNRFKNTIDNLTNISENATNARSRIKDTDFAAETAALSKNQ PROF sec HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH HHHHHHH HHHHHHHHHHHHHH

* *** .*. *.*********.

490 500

STFlic ILQQAGTSVLAQANQVPQNVLSLLR (SEQ ID NO: 1) lUCU HHHHHHHHHHHHHHHHHHHHHH (empirical structure)

PROF_sec HHHHHHHHHHHHH HHHHHHH (predicted secondary structure)

SaFlixl VLQQAGTAILAQANQLPQAVLSLLR (SEQ ID NO: 4)

PROF_sec HHHHHHHHHHHHH HHHHHHH (predicted secondary structure)

.******..******.** ******

[00197] The insertion site, referred to herein as "D3I-ol," in a loop of domain 3 of S. typhimurium FljB (SEQ ID NO: 2) for fusion to an antigen having an isoelectric point greater than about 7.5, is between predicted β-strands VTLA (SEQ ID NO: 131), which is amino acid residues 263-266 of SEQ ID NO: 2, and VVS, which is amino acid residues 293-295 of SEQ ID NO: 2. The insertion site, referred to herein as "D3I-il," in a loop of domain 3 is between β-strands EVNVA (SEQ ID NO: 132), which corresponds to amino acid residues 254-258 of SEQ ID NO: 2, and VTLA (SEQ ID NO: 131 , which corresponds to amino acid residues 263-266 of SEQ ID NO: 2). The insertion site, referred to herein as "D3I-sl," in a loop of domain 3, is between β-strands KYFVTIGG (SEQ ID NO: 133), which corresponds to amino acid residues 234-241 of SEQ ID NO: 2, and EVNVA (SEQ ID NO: 132), which corresponds to amino acid residues 254-258 of SEQ ID NO: 2 in FljB (S. typhimurium, SEQ ID NO: 2). The D3I-ol, D3I-il and D3I-sl insertion sites predicted for S. typhimurium FljB (SEQ ID NO: 2) were loop conformations in the predicted secondary structures of E. coli (SEQ ID NO: 3) and P. aeruginosa (SEQ ID NO: 4) flagellin. Therefore, with reference to the loops in domain 3 of SEQ ID NO: 2, predicted loops in domain 3 and insertion sites, such as D3I-il and D3I-ol, can be selected for E. coli (SEQ ID NO: 3) and P. aeruginosa (SEQ ID NO: 4) flagellin.

[00198] Insertion sites in loops in domain 3 for fusion to antigens with an isoelectric point greater than about 7.0 in E. coli FliC flagellin (SEQ ID NO: 3), can include an D3I-ol site between G274 and A275 of SEQ ID NO: 3 in the loop between β -strand ITF (amino acid residues 267-269 of SEQ ID NO: 3) and β - strand VLTANI (SEQ ID NO: 134, amino acid residues 284-289 of SEQ ID NO: 3). The D3I-il site is between A257 and D258 of SEQ ID NO: 3 in the loop between β - strand ELAKLAIKL (SEQ ID NO: 135, amino acid residues 248-256 of SEQ ID NO: 3) and IEYK (SEQ ID NO: 136, amino acid residues 262-265 of SEQ ID NO: 3). The D3I-sl site was selected between S240 and G241 of SEQ ID NO: 3 in the loop between β-strand KV (residues 238-239 of SEQ ID NO: 3) and β-strand SID (residues 243-245 of SEQ ID NO: 3).

[00199] For P. aeruginosa flagellin (PAOl, SEQ ID NO: 4), the D3I-ol site was selected between Q272 and D273 in the loop between β -strand TVSLA (SEQ ID NO: 137, amino acid residues 262-266 of SEQ ID NO: 4) and β -strand LGITASI (SEQ ID NO: 138, amino acid residues 285-291 of SEQ ID NO: 4). The D3I-il site was between S260 and N261 of SEQ ID NO: 4 in the loop between β -strand SLNFDVTVG (SEQ ID NO: 139, amino acid residues 251-259 of SEQ ID NO: 4) and TVSLA (SEQ ID NO: 137, amino acid residues 262-266 of SEQ ID NO: 4). The D3I-sl site was selected between S245 and G246 of SEQ ID NO: 4 in the loop between β-strand TVFT (SEQ ID NO: 140, residues 238-241 of SEQ ID NO: 4) and β-strand GVT (residues 246-248 of SEQ ID NO: 4).

[00200] Example 2. Fusion proteins of an Influenza B HA antigen that replaces Domain 3 of flagellin

[00201] Compositions that include R3 constructs of flagellin fused to portions of HA have been described (U.S. Patent Application No.: 12/905,584). Previously described R3 fusion proteins include Influenza A HA antigen from PR8 (HI, A/Puerto Rico/8/34, SEQ ID NO: 141), seasonal SI03 (HI , A/Solomon

Island/3/2006, SEQ ID NO: 142), pandemic CA07 (HI , A/California/07/2009, SEQ ID NO: 143), and ΓΝ05 (H5, A/Indonesia/5/2005, SEQ ID NO: 144), which are immunologically potent with minimal side-effects (U.S. Patent Application No.: 12/905,584).

[00202] However, R3 constructs fused to Flu B HA 1-2 portions having a ?-sheet at the bottom of the globular head, such as B/Florida/4/2006 (B/FL) of Yamagata lineage (HL098, SEQ ID NO: 122), were poorly immunogenic in a mouse model (FIG. 2A). In this study, groups of 5-10 BALB/c mice (female, 6-8 weeks old) were immunized twice (day 0 and day 21) subcutaneously (s.c.) or intramuscularly (i.m.) with 10 μg (·) or 3 μg (■) of HL098 (SEQ ID NO: 122) or with 10 μg (A) or 3 μg (▼) of HL199 (SEQ ID NO: 145). The HL098 fusion protein includes a flagellin that lacks domain 3 and replaces it with an HA1-2 (SEQ ID NO: 45), which is referred to herein as an "R3 construct." The HL199 fusion protein includes a flagellin that lacks domains 2 and 3 and replaces domains 2 and 3 with an HAI -2 (SEQ ID NO: 45), which are referred to herein as "R23 constructs." Mice were bled on day 35 after the booster immunization. Serum samples were prepared and stored at -20°C until use. Serum antibody titers were measured by hemaglutination inhibition (HAI) assay or microneutralization (MN) assay.

[00203] The HAI test was used to measure the influenza HA specific antibodies in serum samples (Liu, G., et ah, PLoS ONE 6(6):e20928 (2011); Kendal, A.P., et ah, J. Clin. Microbiol. 18(4):930-934 (1983)). In this assay, immune serum samples, collected at various times after immunization, were treated with receptor destroying enzyme (RDE) II overnight to eliminate non-specific hemagglutination inhibitors. The samples were then heat inactivated (56°C, about 30-45 min), serially diluted starting at about a 1 :5 ratio, and incubated with about 4 HA units (HAU) of influenza B virus for about 30 minutes at room temperature. Turkey red blood cells (0.5%) were incubated with the samples for about 30 to about 60 min at room temperature. HAI titers were measured and reported as the reciprocal of the highest dilution at which hemagglutination is completely inhibited. Ferret post infection anti- influenza B virus (from the Center for Disease Control) is used as reference serum. Serum HAI antibody titers were measured by HAI test against

B/Florida/4/06, and plotted as individual values with geometric mean titers (GMTs). HL098 (SEQ ID NO: 122) elicits no HAI titers.

[00204] F147 buffer, the negative control, contains 10 mM L-histidine, 150 mM NaCl, 5% trehalose, 0.02% polysorbate 80, 0.1 mM EDTA, 0.5% ethanol, and 10 mM Tris at pH 7.2. F147 is a buffer control (♦) and BV (baculovirus) expressed HAO (o, full-length HA) is used as a positive control. As shown in FIG. 2A, none of the mice generated neutralizing antibodies to the original R3 format composition. A modest improvement in immunogenicity was observed with the alternative fusion protein format in which the HAI -2 portion (SEQ ID NO: 45) replaced both D2 and D3 domain, termed R23 (HL199, SEQ ID NO: 145). However, the activity was not sufficient (too low) to advance the composition for use in methods for clinical development. In subsequent testing, R23 (HL199; SEQ ID NO: 145) was used as a base line for comparison to generate new, improved fusion proteins.

[00205] The predicted isoelectric point (pi) of the portion of HA of influenza B (Flu B), for example, the HAI -2 portion of the globular head of Flu B (SEQ ID NO: 45), is between about 8.5 to about 9.5, which is relatively high compared that of HI Influenza A that has a pi of about 7.0, or those of H5 Influenza A that has a pi of about 8.0. The positively charged amino acids in a portion of a globular head of Influenza B HA (e.g., SEQ ID NOs: 8, 9, 45, and 61) may interact with the negatively charged flagellin at neutral pH. As a result, an undesirable charge-charge interaction may distort the fusion protein conformation, which could interfere with TLR5 signaling by the flagellin component of the fusion protein and presentation of the antigen component of the fusion protein. Various efforts to lower the pi of the Flu B globular head, such as introducing negative amino acid residues in non- antigenic regions were unsuccessful. The addition of negatively charged residues in the linking region between HA head and flagellin improved TLR5 signaling and the fusion protein immunogenicity. For example, if 9 amino acid residues (SEQ ID NO: 123) having a negative charge were fused to the C-terminus of the of an HA1-2 portion of influenza B (HL352, SEQ ID NO: 125), which naturally contains two negatively charged residues (E306 and D308), immunogenicity of the Flu B fusion protein was moderately improved (FIG. 2B). This 9 amino acid residue of SEQ ID NO: 123 is native to the naturally occurring HA. In another embodiment, additional negatively charged amino acid residues can be added to the portion of HA or negatively charged amino acid residues can replace naturally occurring amino acid residues at the carboxy -terminus of the HA.

[00206] In this study, serum was evaluated in the microneutralization (MN) test ( owe, T., et al, J. Clin. Microbiol 37:937-943 (1999); Song, L., et al. PLoS ONE 3:e2257 (2008)). Briefly, groups of 10 BALB/c mice were treated s.c. with fusion protein compositions containing HL199 (·, SEQ ID NO: 145) or HL352 (▼, SEQ ID NO: 126) on days 0 and 21 , and bled on day 35. Serum samples were treated with receptor destroying enzyme II, serially diluted in duplicate, starting at a ratio of about 1 :10 and co-cultivated with 100 TQD_SO (tissue culture infective dose, TCID) of influenza B virus for about 1.5 hr. Madin-Darby Canine Kidney (MDCK) cells (about 4 x ¹⁰⁴ /well) were then added and incubated for about 20 hours at about 37°C. Cells were washed, fixed, air-dried and incubated with a monoclonal anti- influenza B nucleoprotein antibody (primary) and goat anti-mouse Fey specific IgG:HRP (secondary). Serum antibody titers were measured by MN test against B/Florida/4/06 virus, and plotted individually with GMTs (horizontal lines, FIG. 2B). Control groups include the formulation buffer (+, F147, negative) and BV (baculovirus) expressed HA0 (*, full-length HA, positive). The negative charges in the linker fusing the HA portion to flagellin may alleviate the charge-charge interactions and allow flagellin to bind to TLR5 more efficiently, which results in adaptive immune responses.

[00207] The ability of fusion proteins described herein to stimulate the innate arm of the immune system, i.e., the TLR5 bioactivity, was assessed by an in vivo TLR5 assay The assay consists of an in-life phase and an in vitro serum cytokine rise readout. In general, groups of BALB/c mice (5 per group) were injected with vehicle control (naive) or test fusion protein compositions at about 1 μg each. Three hours later, mice were bled and the serum was prepared and analyzed for inflammatory cytokines using a mouse cytometric bead array (CBA) from BD (BD Biosciences, San Jose, CA), and run on a flow cytometer.

[00208] For the cytokine assessment, only IL-6 and TNF-a displayed a range of responses that are useful for characterizing fusion protein composition candidates while others either produced only baseline levels at the dose and time point chosen, or produced plateau levels which are not useful for differentiation of composition potency to stimulate TLR5 activation (data not shown). Using naive serum levels, as well as those from low potency fusion protein (STF2.HA1-2, B/Florida/4/2006, SEQ ID NO: 146) that generated low micro-neutralization (MN) titers, thus the minimum innate immune stimulation required for an immune response, 4 zones of activity were established. These 4 zones of activity levels are depicted in FIGs. 3 A and 3B. The lowest level of cytokine activity, which is designated as inactive, consists of levels less than about or equal to the mean plus 3 standard deviations (SD) of that detected in naive mice. The next level, which is designated as low, consists of levels from greater than about the mean plus 3 SD of the naive and less than about or equal to the mean plus 1 SD of low potency fusion protein (SEQ ID NO: 146). The next zone, which is designated as moderate, consists of values greater than about the mean plus 1 SD of low potency fusion protein composition but less than about or equal to the mean plus 3 SD of high potency fusion protein composition (HL185, STF2R3.HA1-2, an R3 construct) (A/California/07/2009, CA07), SEQ ID NO: 143). Finally, the fourth zone, designated as high, is for values greater than about the mean plus 3 Standard deviations (SD) of high potency composition.

[00209] In vivo TLR5 bioassays were employed to evaluate charge-charge interactions that may influence TLR5 binding activity. The addition of a 9 amino acid negatively charged linker (SEQ ID NO: 123) to a fusion protein that includes an HA 1-2 portion of Influenza B/Flori da/4/2006 to generate the fusion protein referred to as HL352 (SEQ ID NO: 125) improved cytokine responses, as shown by elevated IL6 and TNF-a responses (FIGs. 3A and 3B).

[00210] Similar results were obtained for a fusion protein employing sequences from B/Brisbane/60/2008 (B/BR) as a prototypical composition for the Victoria lineage (FIGs. 4 A and 4B). FIGs. 4 A and 4B show the Balb/c mice were left naive (♦) or injected with the indicated fusion protein composition for the B/BR (Victoria lineage), at a 1 μg dose: R3.HA1-2 B (T, HL169, SEQ ID NO: 147) and R3.HA1-2 B with a 9 amino acid linker B (■, HL483, SEQ ID NO: 148). The R3.HA1 (·, HL185, SEQ ID NO: 143) fusion protein was included as a positive control. At 3 hours, mice were bled to generate serum. Cytokine levels were quantified using a mouse inflammation cytometric bead array (BD Biosciences, San Jose, CA). Dotted lines in FIGs. 4A and 4B indicate the activity levels discussed above.

[00211] An additional three negatively charged amino acid residues were added to the carboxy- terminus of the portion of HA in the fusion proteins HL610 (SEQ ID NO: 149) for B/FL and HL611 for B/BR (SEQ ID NO: 150) for a total of 5 negatively charged amino acid residue extension. The three negatively charged residues are native to the naturally occurring HA sequence. For example, the three negatively charged amino acid residues substitutions can be K298E, S300D and I304D in HA B/FL of SEQ ID NO: 124 or 299E, S301D and I305D in HA

B/BR60 of SEQ ID NO: 264. The three negatively charged amino acid residues are not necessary a continuous stretch of amino acid residues in the naturally occurring HA protein. The negatively charged amino acid residues may reduce charge-charge interaction between the positively charged portion of the HA globular head and the negatively charged flagellin, to thereby minimize charge-charge intramolecular interactions between the antigen and flagellin component of the fusion proteins.

[00212] The in vivo TLR5 bioassay was performed as above with Balb/c mice were left na^'ive (■) or injected with the indicated fusion protein composition at a 1 μg dose: B/FL (Yamagata lineage) R3.HA1-2 B with a 9 amino acid linker of SEQ ID NO: 123 (■, HL352, SEQ ID NO: 125) and the same format with additional negative charges (described supra, K298E, S300D and I304D in HA B/FL in SEQ ID NO: 125 and K299E, S301D and 1305 in HA B/BR60 in SEQ ID NO: 264 introduced in the linker (·, HL610, SEQ ID NO: 149) and B/BR (Yamagata lineage) R3.HA1-2 B with a 9 amino acid linker (SEQ ID NO: 123) (A , HL169, SEQ ID NO: 147) and the same format with additional negatively charged amino acids introduced in the linker (*, HL611, SEQ ID NO: 150). R3.HA1 (·, HL185, SEQ ID NO: 143) was included as a positive control. Dotted lines in FIGs. 5 A and 5B indicate thresholds of cytokines levels correlated with immunogenicity as, described above. Constructs HL610 (SEQ ID NO: 149) and HL61 1 (SEQ ID NO: 150) improved the cytokine profiles compared to constructs HL352 (SEQ ID NO: 125) and HL169 (SEQ ID NO: 147) (FIGs. 5 A and 5B).

[00213] In addition to the in vivo TLR5 cytokine assay that was used to assess the functionality of flagellin moiety, a neutralization inhibition assay (NIA) was employed to evaluate the potency of the HA moiety. NIA was used to measure the ability of a fusion protein composition to compete with virus for binding to neutralizing antibodies present in sheep hyper-immune serum (available from CBER, Center for Biologies Evaluation and Research in FDA), similar to that used for the potency release assay (SRID, Single Radial Immunodiffusion) of a commercial influenza composition (TIV, Trivalent Influenza Vaccine (TIV) containing two different inactivated influenza type A strains and one inactive influenza type B strain, e.g. FLUVIRIN^®), thereby assessing the integrity of neutralizing epitopes on the fusion proteins described herein.

[00214] In the NIA assay, higher potency of the fusion proteins was demonstrated by increased depletion of the neutralizing antibodies in the sheep hyper-immune serum, which allows the test virus to infect MDCK cells. The sheep hyper-immune serum raised against the test Flu B influenza virus (sheep anti-HA of B/Florida/4/06 serum (FDA) or sheep anti-HA of B/Brisbane/60/08 serum (FDA)) was pre- incubated with serially diluted test articles at about a 1 : 1 ratio for about 90 minutes. 50 TCID₅₀ B/Florida/4/06 virus or 50 TCID₅₀ B/Brisbane/60/08 virus was then added and allowed to incubate for about 1 hour at 37 °C prior to addition of MDCK cells. Following an incubation of about 18 hours to about 20 hours of the mixture with MDCK cells at about 37°C, the cells were fixed with a cold solution of 80% acetone in DPBS (1.47 mM KH2P04, 2.67 mM KC1, 138 mM NaCl, 8.06 mM Na2HPD4.7H20, Invitrogen Cat. No. 14190-250).

[00215] Intracellular influenza virus was quantified in an ELISA (enzyme-linked immunosorbent assay) format using influenza B NP-specific (nuclear protein- specific) mAbs (monoclonal antibodies) as primary antibodies and horseradish peroxidase (HRP) conjugated goat anti-mouse IgG (Jackson ImmunoResearch, Cat. no. 1 15-035-008). After color development with TMB (3, 3'-5,5'- tetramethylbenzidine) substrate (ThermoScientific, Cat. no. 34028), the virus was quantified by measuring OD₄₅₀. NIA curves were generated by Log-Logit fit model using Softmax software 5.2. The greater the integrity of the neutralizing epitopes in the composition the higher infectivity of the virus and the higher the signal in the NIA assay.

[00216] To complement the TLR5 assay, NIA assays were performed to determine the integrity of HA antigens in the Flu B compositions that contain negative linkers (SEQ ID NOs: 149 for HL610 and 150 for HL611), described supra. As shown in FIG. 6, B/Florida/4/06 R3 compositions of the fusion proteins HL352 (SEQ ID NO: 125) or HL610 (SEQ ID NO: 149) exhibit comparable ability to inhibit antibody neutralization. As shown in FIG. 7, B/Brisbane/60/08 R3 compositions of the fusion proteins HL169 (SEQ ID NO: 147), HL483 (SEQ ID NO: 148), and HL611 (SEQ ID NO: 150) also show similar integrity of the neutralizing epitopes. These results indicate that influenza B portions of HA globular heads fused to a loop of domain 3 of flagellin are folded similarly and present epitopes in a conformation similar to that of the wild type virus. Given this similarity, the difference in the ability of fusion proteins described herein, such as FluB antigens fused to a loop of domain 3 with or without a 9 amino acid extension (SEQ ID NO: 123) or with or without an additional three negatively charged amino acid residues corresponding to amino acid residues of the naturally occurring HA, to trigger TLR5 signaling becomes critical to determine the fusion protein potency.

[00217] Example 3. Fusion proteins with HA antigen inserted into loop regions of domains 1, 2, or 3 of flagellin to shift pi to target influenza B

[00218] Although negative charges in the linker regions improved the TLR5 signaling, the immunogenicity of these fusion proteins (SEQ ID NOs: 149 and 150) remained suboptimal, especially for Victoria lineages, such as B/BR. The negative linker may only locally adjust the orientation of HA head with respect to flagellin, which may not be sufficient to allow the formation of 2:2 heterodimers that are critical to initiate a TLR5 signaling cascade. As described below, a portion of an HA globular head was positioned further away from domain 1 of flagellin, the TLR5 binding domain. For example, as depicted in FIG. 14, domain 1 of S, typhurium FliC flagellin is between amino acid residues 47-176 of SEQ ID NO: 2 (amino- domain 0) and amino acid residues 415-464 of SEQ ID NO: 2 (carboxy-domain 0). Fusion protein constructs (HL656, SEQ ID NO: 151, D3Ins of B/FL and HL657, SEQ ID NO: 128, D3Ins of B/BR) were created by fusion of a portion of the globular head of HA antigen into domain 3 of flagellin, which are referred to as domain 3 insertions ("D3Ins").

[00219] As shown in FIGs. 8A and 8B, more favorable cytokine profiles were observed with the D3Ins fusion protein as assessed by in vivo TLR5 assays. As described above, Balb/c mice were left naive (♦) or injected with the indicated composition candidate, at about a 1 μg dose: B/FL (Yamagata lineage) R3.HA B (■, HL352, SEQ ID NO: 125) and D3Ins.HA B ( A, HL656, SEQ ID NO: 151), B/BR (Victoria lineage) R3.HA B (A , HL169, SEQ ID NO: 147) and D3Ins.HA B (T , HL657, SEQ ID NO: 128). R3.HA1 (·, HL185, SEQ ID NO: 143) was included as a positive control. Fusion proteins generated by removal of domain 3 of flagellin and replacement with a portion of an HA globular head, referred to as "R3" fusion proteins (e.g., R3.HAB) were made employing previously described methods (see, for example, U.S. Application No: 12/905,584). NIA assay did detect an

appreciable difference between D3Ins fusion protein and the R3 fusion protein, indicating the antigen retained similar integrity in both formats (FIGs. 9A-9C). The NIA assay assesses the functionality of the portion of the globular head of HA.

[00220] The NIA assay was performed as described above using about a 1 : 1 ratio of serially diluted test articles B/Florida/4/06 (o, HL352, SEQ ID NO: 125 and□, HL656, SEQ ID NO: 151) (FIG. 9A); B/Brisbane/60/08 (□, HL611, SEQ ID NO: 150 and o, HL657, SEQ ID NO: 128) (FIG. 9B); or B/Wisconsin/1/10 (▲, HL724, SEQ ID NO: 152 and o, HL772, SEQ ID NO: 126) (FIG. 9C) mixed with sheep anti-HA of B/Florida/4/06 serum (A), B/Brisbane/60/08 (B), or B/Wisconsin 1/10 (C). It is believed that TLR5 signaling plays a more critical role than neutralization inhibition in eliciting protective immune responses. When mice were immunized with a D3Ins fusion protein, D3Ins format compositions including HL610 (SEQ ID NO: 149) (o, R3 fusion protein with 9 amino acid extension and the addition of negatively charged amino acids (SEQ ID NO: 153)) and HL656 (SEQ ID NO: 151) (o, D3Ins with 9 amino acid extension (SEQ ID NO: 123)) elicited stronger immune responses than other fusion proteins, such as HL352 (SEQ ID NO: 125) (·, R3 with the 9 amino acid extension of SEQ ID NO: 123), (FIG. 10). HA352 (SEQ ID NO: 125) is a R3 fusion protein that includes an HA1-2 portion of the globular head of influenza B FL4. Serum antibody titers in FIG. 10 were measured by MN test using B/Florida/4/06 virus antigen. Data are shown as titers of individual mice with GMTs and seroconversion rates (percent (%) of mice with a MN titer > 80) above each group. These data show that the D3Ins format (HL656, SEQ ID NO: 151) was superior to the R3 format for Flu B (HL352, SEQ ID NO: 125).

[00221] The immunogenicity of D3Ins fusion proteins listed in Table 3 were evaluated using B/Wisconsin/1/2010, a recently circulating Yamagata lineage strain. In an immunogenicity study, groups of 6 BALB/c mice were immunized with 3 different formats of fusion proteins that include flagellin and portions of the globular head of the HA of Influenza B at a 6 μg dose delivered s.c. in two injections on day 0 and day 21 with the Flu B compositions: R3 B/WI (·, HL719, SEQ ID NO: 154, R3 with negative linker), R3 B/WI (■, HL724, SEQ ID NO: 152, R3 with additional negative charges in the linker), D3Ins B WI (A, HL772, SEQ ID NO: 126), or F 147 buffer (·). Mice were bled on day 35 for HAI titers. Controls included a group immunized with F147 buffer (10 mM L-histidine, 150 mM NaCl, 5% trehalose, 0.02% polysorbate 80, 0.1 mM EDTA, 0.5% ethanol, 10 mM Tris, pH 7.2). All fusion proteins were prepared in F 147 buffer. Sera were tested with HAI assays using extracted B/Wisconsin/1/2010 virus antigen. Data are shown as titers of individual mice with GMTs. Statistical differences were determined in 1-way ANOVA/Tukey test with *, /? < 0.05, ***,/? < 0.001.

[00222] As illustrated in FIG. 1 1, the highest titers in the monovalent groups were seen with D3Ins fusion protein (HL772, A , SEQ ID NO: 126) followed by an R3 fusion protein HL724 (■, SEQ ID NO: 152, B/WI counterpart of B/FL HL610, SEQ ID NO: 149) and HL719 (·, SEQ ID NO: 154, B/WI counterpart of B/FL HL352, SEQ ID NO: 125). Titers of both HL724 (R3 format) and HL772 (D3Ins format) groups were significantly higher than the F147 control. This is consistent with the same formats of B Florida and B Brisbane where the D3Ins was highest, followed by R3 formats (FIG. 10 for B/FL, FIG. 12 for B Victoria lineage, including

Brisbane, Hong Kong and Bangladesh). Thus, as predicted by TLR5 data, the D3Ins format improved immunogenicity to the B/WI strain compared to the R3 formats for fusion proteins in mice. A positive immunization control was not included, because the immunogen for B/Wisconsin/1/10 was unavailable at the time of the analysis.

[00223] HAI testing of immune sera from a mouse immunogenicity study comparing R3 and D3Ins formats of B Brisbane-like strains was performed. The same modifications were made to the HA antigen (SEQ ID NOS: 95-97), and these antigens were employed in the R3 and D3Ins formats. Groups of 6 BALB/c were immunized s.c. on days 0 and 21 with 6 μg of fusion protein R3 compositions (HL61 1. SEQ ID NO: 150; HL753, SEQ ID NO: 155; HL742, SEQ ID NO: 156) or fusion protein D3Ins compositions of B/Brisbane/60/08-like viruses

(B/Brisbane/60/08 or B/BR, HL657, SEQ ID NO: 128; B/Hong Kong/259/10 or B/HK, HL774, SEQ ID NO: 157; and B/Bangladesh/5945/09 or B/BD, HL787, SEQ ID NO: 158), and bled on day 35. Mice in control groups received 15 μg

FLUVIRIN^® or F 147 buffer. Neutralizing antibody titers of the serum samples were measured by HAI test using ether extracted B/Brisbane/60/08, B/Hong

Kong/259/10, or B/Bangladesh/5945/09 vims, and expressed as GMTs. Statistical differences were determined in 2-way ANOVA/Tukey tests with *, p < 0.05, **,p < 0.01.

[00224] Results of this study indicated that the D3Ins format of

B/Brisbane/60/2008-like fusion protein (B/Brisbane/60/08, B/BR SEQ ID NO: 128; B/Hong Kong/259/10, B/HK SEQ ID NO: 157 and B/Bangladesh/5495/09, B/BD SEQ ID NO: 158) elicited significantly higher HAI titers than the strain-matched R3 format measured using 3 different B/Brisbane-like viruses (FIG. 12). D3Ins B/BD (HL787, SEQ ID NO: 158) elicited HAI geometric mean titers comparable to FLUVIRIN^® containing B Brisbane, particularly when B/Hong Kong and

B/Bangladesh viruses were used as HAI antigens. These data show that the D3Ins fusion proteins consistently have higher hemagglutination inhibition than R3 fusion proteins across multiple influenza B strains and within a cluster of antigenically similar viruses. Fusion proteins that include portions of the globular head of influenza B fused to a loop of domain 3 of flagellin improved immunogenicity compared to HA portions fused to R3 constructs.

[00225] Example 4. Screening the insertion variant as a Flu B fusion protein

[00226] The initial design of the insertion fusion proteins targeted the D3 domain as the insertion site. Three insertion sites (D3I-ol, D3I-il and D3I-sl) in loops of domain 3 were selected for fusion of portions of HA antigens (see, for example, FIG. 29 and 30). In addition to the B/FL fusion protein, B/Wisconsin/1/2010 (B/WI, SEQ ID NO: 263), a currently circulating B Yamagata lineage was chosen to develop seasonal Flu B fusion protein for use in methods of stimulating immune responses. B/BR remains as another candidate in fusion protein development representing another circulating B Victoria lineage Flu B strain. The three constructs of B/WI (HL772, D3I-ol, SEQ ID NO: 126; HL849, D3I-il, SEQ ID NO: 159; and HL848, D3I-sl, SEQ ID NO: 160) were cloned into the PET24a expression vector, and proteins were purified from E. coli cell culture using standard fermentation as previously described (Song, L., et al, PLoS One 3:e2257 (2008); Song, L., et al, Vaccine 27:5875-5884 (2009)).

[00227] The TNF-oc and IL-6 cytokine levels of B/WI constructs are depicted in FIGs. 13A and 13B. The R3 format of HI CA07 R3.HA1 (HL185, also referred to as "VAX128," SEQ ID NO: 143) fusion protein was used as a positive control. The CA07 vaccine R3.HA1 (HL185, SEQ ID NO: 143) fusion protein has been clinically proven to be safe and immunogenic in humans (Taylor, D., et al, Vaccine 30:5761- 5769 (2012) and a strong stimulator of TLR5. The positive controls elicited high levels of TNF-a and IL-6.

[00228] The fusion proteins that include an HA antigen fused to a loop of domain 3, B Wisconsin (Yamagata lineage), including HL772 (D3I-ol, SEQ ID NO: 126) and HL849 (D3I-il, SEQ ID NO: 159), elicited comparable IL-6 levels in the "High" zone with HL848 (D3I-sl, SEQ ID NO: 160) the lowest in the "Moderate" zone. The TNF-a levels for HL772 (D3I-ol, SEQ ID NO: 126) and HL849 (D3I-il , SEQ ID NO: 159) were lower compared to a positive control but were well above the minimum threshold required to achieve proper immune responses. The fusion protein HL848 (D3I-sl, SEQ ID NO: 160) had TLR5 activity in the "Inactive" zone, indicating the D3I-sl format may not elicit a strong immune response (FIGs. 13 A and 13B). These data show that fusion proteins that include an antigen fused to a loop of domain 3 of flagellin have fewer side effects (are less reactogenic) than fusion proteins that include antigens fused to R3 constructs of flagellin, yet elicit suitable cytokine activity that is critical for immune responses. See Tables 1, 3 and 4 for a list of fusion proteins and sites of insertion in flagellin.

[00229] To assess for a fusion protein that maximizes TLR5 signaling and immunogenicity with minimal side effects, loops in domain 2 and domain 1 flagellin, in addition to loops in domain 3 of flagellin, were selected as possible sites of fusion to antigens. The insertion (fusion) sites in flagellin were located in the loop regions so the HA antigen could be inserted without disturbing the overall flagellin structure conformation, including the tertiary structure of domain 3 of flagellin as depicted in FIGs. 14 and 29. The primary amino acid sequence of flagellin indicating the amino acid residues of flagellin between which portions of the globular head of HA from influenza B were fused is depicted in FIGs. 29 and 30. The predicted loops in domains of flagellin is based on a comparison of the known tertiary structure of S. typhimuium FliC flagellin, as discussed supra. Exemplary fusion proteins of the invention include the Yamagata lineage B/Wisconsin/1/2010 listed in Table 3, and Victoria lineage B/Brisbane/60/2008 listed in Table 4.

[00230] Table 3. Flu B Yamagata Lineage (B/Wisconsin/1/2010) Fusion Proteins. The fusion protein includes a portion of the globular head of HA (SEQ ID NO: 50) fused to a loop of a domain of S. typhimuium FljB flagellin (SEQ ID NO: 2 or 127).

[00231] TABLE 3

Fusion Protein SEQ ID NO: Format Insertion Site

HL772 126 D3I-ol 276-277 in SEQ ID NO: 127

HL656 151 D3I-ol 276-277 in SEQ ID NO: 127

HL849 159 D3I-il 259-260 in SEQ ID NO: 2 HL848 160 D3I-sl 244-245 in SEQ ID NO: 2

HL825 161 D2I-ol 187-188 in SEQ ID NO: 2

HL826 162 D2I-o2 178-179 in SEQ ID NO: 2

HL827 163 D2I-o3 326-327 in SEQ ID NO: 2

HL828 164 Dl l-ol 102-103 in SEQ ID NO: 2

HL850 165 D2I-il 366-367 in SEQ ID NO: 2

HL888 169 D3I-o2 230-231 in SEQ ID NO: 2

HL890 170 D2I-cl 308-309 in SEQ ID NO: 2

HL892 171 D2I-i2 388-389 in SEQ ID NO: 2

[00232] Table 4. Flu B Victoria Lineage (B/Brisbane60/2008) Fusion Proteins. The fusion proteins include a portion of the globular head of HA (SEQ ID NO: 48) fused to a domain of S. typhimuium FljB flagellin (SEQ ID NO: 2, 127, or 270).

[00233] TABLE 4

Fusion Protein SEQ ID Format Insertion Site

NO:

HL657 128 D3I-ol 276-277 in SEQ ID NO: 127

HL733 166 D2I-ol 187-188 in SEQ ID NO: 270

HL856 167 D2I-ol 187-188 in SEQ ID NO: 2

HL857 168 D2I-o2 178-179 in SEQ ID NO: 2

HL858 172 D2I-o3 326-327 in SEQ ID NO: 2

HL860 173 D3I-sl 244-245 in SEQ ID NO: 2

HL889 174 D3I-o2 230-231 in SEQ ID NO: 2

HL891 175 D2I-cl 308-309 in SEQ ID NO: 2 HL893 176 D2I-i2 388-389 in SEQ ID NO: 2

HL861 180 D3I-il 259-260 in SEQ ID NO: 2

HL862 181 D2I-il 366-367 in SEQ ID NO: 2

[00234] The fusion protein listed in Tables 3 and 4 were evaluated using a TL 5 assay, as described supra. Comparison of fusion proteins with insertions into domain 3 of flagellin (D3I) to insertions into domain 2 of flagellin (D2I) are shown in FIGs. 15A-15B and 16A-16B. The fusion protein constructs used were B

Wisconsin (Yamagata lineage) D3I-ol (HL772, SEQ ID NO: 126), D2I-ol (HL825, SEQ ID NO: 161), D2I-o2 (HL826, SEQ ID NO: 162), D2I-o3 (HL827, SEQ ID NO: 163), Dll-ol (HL828, SEQ ID NO: 164) and D2I-il (HL850, SEQ ID NO: 165) and B Brisbane (Victoria lineage) D3I-ol (HL657, SEQ ID NO: 128), D2I-ol (HL733, SEQ ID NO: 166), D2I-ol (HL856, SEQ ID NO: 167) and D2I-o2 (HL857, SEQ ID NO: 168) (FIGs. 15A-15B). The fusion protein constructs used were B Wisconsin (Yamagata lineage, FIG. 16A) D3I-ol (HL772, SEQ ID NO: 126), D3I- o2 (HL888, SEQ ID NO: 169), D2I-cl (HL890, SEQ ID NO: 170), D2I-i2 (HL892, SEQ ID NO: 171) and B Brisbane (Victoria lineage, FIG. 16B) D3I-ol (HL657, SEQ ID NO: 128), D2I-o3 (HL858, SEQ ID NO: 172), D3I-sl (HL860, SEQ ID NO: 173), D3I-o2 (HL889, SEQ ID NO: 174), D2I-cl (HL891, SEQ ID NO: 175) and D2I-i2 (HL893, SEQ ID NO: 176). R3.HA1 (HL185, SEQ ID NO: 143) was included as a positive control.

[00235] These data show that fusion proteins that include antigens fused to loops in domain 3 of flagellin ("D3 insertion constructs, " "D3I," "D3Ins," "D3Ins fusion proteins") had enhanced TLR5 activity compared to fusion proteins that include an antigen fused to loops in domain 2 of flagellin ("D2 insertion constructs"). Fusion proteins that include an antigen fused to a loop in domain 1 of flagellin ("Dl insertion constructs") were not active. Among the D3Ins fusion proteins, D3I-ol (HL772, SEQ ID NO: 21) and D3I-il (HL849, SEQ ID NO: 22) were more active than D3I-sl (HL848, SEQ ID NO: 23) (FIGs. 14-16).

[00236] Antibodies generated to fusion proteins having portions of the globular head fused to domain 3, domain 2, or domain 1 were compared for their ability to inhibit virus neutralization in an NIA assay. The NIA assay data for antibodies generated to fusion proteins include (B/Wisconsin/1/2010): D3Ins (o, HL772, D3I- ol , SEQ ID NO: 126), various D2Ins (Δ, HL825, D2I-ol, SEQ ID NO: 161 ; 0, HL826, D2I-o2, SEQ ID NO: 162; ·, HL827, D2I-o3, SEQ ID NO: 163), and Dllns (■, HL828, Dll-ol, SEQ ID NO: 164), in comparison to R3 format (A, HL724, SEQ ID NO: 152) and are shown in FIG. 17. The NIA assay data for D3I insertion fusion proteins in comparison to the D2I insertion fusion proteins are shown in FIG. 18. The portion of the globular head fused to a loop of domain 2 or a loop of domain 3 is has the same amino acid sequence (SEQ ID NO: 50).

[00237] FIG. 18A shows NIA assay data for B/Wisconsin/1/2010 D3I insertion variants: D3I-ol (o, HL772, SEQ ID NO: 126), D3I-sl (Δ, HL848, SEQ ID NO: 160), D3I-il (0, HL849, SEQ ID NO: 159), and D3I (A , HL864, SEQ ID NO: 177), in comparison to the D2I-il (·, HL850, SEQ ID NO: 165) and D2I (■, HL854, SEQ ID NO: 179) fusion proteins. FIG. 18A shows the NIA assay data for

B/Wisconsin/1/2010 D3I variants: D3I-ol (o, HL772, SEQ ID NO: 126), D3I-o2 (□, HL888, SEQ ID NO: 169), and D2I-cl (Δ, HL890, SEQ ID NO: 170), in

comparison to D2I-i2 (0, HL892, SEQ ID NO: 171).

[00238] Relatively higher NIA titers were observed with D3 insertion fusion proteins compared to the R3 format as demonstrated by the comparison of HL772 (o, D3Ins, SEQ ID NO: 126) to HL724 ( A, R3, SEQ ID NO: 152) (FIG. 17), indicative of slightly better antigen presentation for D3 insertion format than R3 format. NIA titers indicate that fusion of an antigen to flagellin can deplete serum neutralizing antibodies, which may not correlate with the ability of the fusion protein to generate a protective immune response. The ability to generate a protective immune response includes a balance of the ability of the fusion protein to activate TLR5 and stimulate a sufficient immune response with minimum side effects.

[00239] Advantages of fusing antigens to a loop of domain 3 of flagellin by inducing key cytokine secretion in response to immunization is depicted in FIGs. 9A-9C. Taken together, the cytokine and NIA results indicated that the D3 Insertion format of flagellin was a superior format for fusing portion of the globular head of influenza B compared to the R3 format. Furthermore, D3 insertion fusion proteins exhibited higher neutralizating antibody titers compared to D2 insertion fusion proteins. Consistent with the cytokine assay, Dl insertion fusion proteins were inactive in NIA assay. Among three D3 insertion fusion proteins, D3I-ol and D3I- il again demonstrated a greater ability to neutralize viral infection than D3I-sl . The NIA data is consistent with the cytokine activation to assess TLR5 activation.

[00240] The NIA was employed to evaluate fusion proteins that include portions of the globular head of influenza B for the Flu B Victoria lineage using

B/Brisbane/60/2008 as a prototype. The NIA assay data for D3I fusion proteins: D3I-ol (o, HL657, SEQ ID NO: 128) and D3I-il (B, HL861, SEQ ID NO: 180), compared to the R3 (□, HL611, SEQ ID NO: 150) fusion protein are shown in FIG. 19A. FIG. 19B shows the NIA assay data for B/Brisbane/60/2008 R3 (□, HL611, SEQ ID NO: 150) fusion protein in comparison to other fusion proteins in which portions of the globular head are fused to other regions of domain 3, or domain 2: D3I-ol (o, HL657, SEQ ID NO: 128), D2I-ol (Δ, HL733, SEQ ID NO: 166), D2I- ol (■, HL856, SEQ ID NO: 167), D2I-o2 (▲, HL857, SEQ ID NO: 168), and D2I- il (♦, HL862, SEQ ID NO: 181). The NIA assay data for D3I fusion proteins: D3I- ol (o, HL657, SEQ ID NO: 128) and D2I-o3 (□, HL858, SEQ ID NO: 172), in comparison to D2I fusion proteins: D3I-sl (Δ, HL860, SEQ ID NO: 173), D3I-o2 (HL889, SEQ ID NO: 174), D2I-cl (HL891, SEQ ID NO:T75), D2I-i2 (HL893, SEQ ID NO: 176), and D2I-il (HL862, SEQ ID NO: 181), are shown in FIG. 20.

[00241] As shown in FIG. 19A, the D3Ins fusion proteins were comparable to the R3 fusion protein in neutralization inhibition, which indicated similar integrity in both fusion protein formats. When the in vivo cytokine data for the TLR5 test was evaluated, it was determined that the R3 format fusion protein of

B/Brisbane/60/2008 was a poor trigger of cytokine release, which is believed to be a key event to elicit an adaptive immune response. Therefore, the D3Ins fusion protein is an exemplary format for Flu B Victoria lineage.

[00242] As shown in FIG. 19B, the D2Ins fusion proteins was similar, generating activity slightly lower as compared to D3Ins fusion proteins. However, the TLR5 activity of the D2Ins fusion protein constructs was significantly lower than D3Ins fusion proteins (FIGs. 15 A, 15B, and 16B). Consequently, D3Ins fusion proteins, such as HL657 (D3I-ol, SEQ ID NO: 128) and HL861 (D3I-il, SEQ ID NO: 180), are believed to be better for methods of stimulating an immune response.

[00243] In conclusion, fusion of an antigen with an isoelectric point greater than about 7.0 to a loop of domain 3 of flagellin ("D3Ins fusion proteins") can significantly improve TLR5 bioactivity compared to the R3 format. In the NIA functional assay, D3Ins fusion proteins also demonstrated equivalent or greater activity to compete with virus for neutralizing antibodies. D3 insertion format is suitable, for example, as a fusion protein format for use in fusing portions of the globular head of the HA influenza B to flagellin, and is superior to either R3 or D2Ins formats for use with antigens of HA of influenza B, or influenza A antigens that have an isoelecric point greater than about 7.0, such as about 7.5, about 8.0 or about 8.5., about 9.0, about 9.5, about 10.5 or about 11.0.

[00244] Example 5. Immunogenicity of fusion proteins with HA antigen inserted into loop regions of domain 3 of flagellin to shift pi to target influenza B in a mouse model

[00245] The immunogenicity of three different fusion proteins inserted into three different regions of loops of domain 3 was evaluated in mice. These fusion proteins were D3I-ol, HL772, (SEQ ID NO: 126), D3I-il, HL849 (SEQ ID NO: 159), and D2I-il, HL850 (SEQ ID NO: 165) of B/Wisconsin/1/2010 (B/WI). Groups of 8 BALB/c mice were treated i.m. with 2 dose levels (3 μg and 12 g) of D3I-ol B/WI (HL772, SEQ ID NO: 126), D3I-il B/WI (HL849, SEQ ID NO: 159), and D2I- ilB/WI (HL850, SEQ ID NO: 165) fusion protein on days 0 and 21. All animals were bled on day 35. Serum HAI antibodies were measured by HAI test using ether-extracted B/Wisconsin/1/2010 virus, and horizontal lines represented geometric mean titers (GMTs) (GMT values above the lines above) in FIG. 21 A. Seroconversion rates (% of mice showing an HAI > 40) were also given in FIG. 21 A as percentages. Statistical differences were determined against F147 buffer control group in ANOVA tests with ***, < 0.001.

[00246] As shown in FIG. 21 A, D3I-ol B/WI elicited about 2-fold higher HAI antibody titers compared to D3I-il B/WI. Inactivated virus antigen of B/Hubei- Wujiagang/158/2009 (a B/Wisconsin l/2010-like virus obtained from CBER, HA HU at 6 μg in FIG. 21 A) was used as a positive control and also elicited significant HAI titers. In contrast, D2I-il B/WI (HL850) and baculovirus expressed rHAO of Protein Sciences induced significantly lower HAI titers. Buffer F147 was used as a negative control (FIG. 21 A). Therefore, the immunogenicity of three

B/Wisconsin/1/2010 fusion proteins ranked as D3I-ol > D3I-il > D2I-il, consistent with results from cytokine tests and NIA assays.

[00247] The D3Ins fusion proteins of B/Sichuan/379/1999 (B Yamagata lineage included in 2001/2002 TIV and B/Florida/4/2006 (in 2008/2009 TIV) were also evaluated for immunogenicity as shown in FIG. 2 IB. Groups of 10 mice were immunized s.c. with 3 different doses (0.05 μg, 0.5 μg and 5 μg) of

B/Sichuan 379/1999 D3I-ol (HL863, SEQ ID NO: 182) or D3I-il(HL903, SEQ ID NO: 183) fusion protein on days 0 and 21. As shown in FIG. 21B, the D3Ins fusion proteins of B/Sichuan/379/1999 were immunogenic with no statistically significant difference between ol or il formats. In general, the D3Ins is more immunogenic than the D2Ins constructs.

[00248] Example 6. D3Ins fusion protein format

[00249] In order to confirm the general applicability of the D3Ins format of flagellin for use in generating fusion proteins of portions of the globular head of influenza B, fusion proteins for three currently circulating strains in the Yamagata lineage were evaluated in a mouse immunogenicity study. In this study, treatment of groups of 8 BALB/c mice immunized with 2 doses (3 μg and 12 μg) of fusion protein prepared with HA sequence from B/Wisconsin 1/2010 (D3Ins B/WI, HL772, SEQ ID NO: 126); B/Hubei-Wujiagang/158/2009 (D3Ins B/HU HL869, SEQ ID NO: 184); and B/Texas/6/2011 (D3Ins B/TX, HL871, SEQ ID NO: 185). HL869 includes a portion of HA of SEQ ID NO: 58. HL871 includes a portion of HA of SEQ ID NO: 60.

[00250] As shown in FIG. 22, the D3Ins fusion proteins of the three strains elicited significant HAI titers (GMTs = 71 - 283 in 3 μg dose groups; 273-719 in 12 μg dose groups). Furthermore, D3Ins B/WI at a 3 μg dose elicited a 12-fold higher geometric mean HAI titer (GMT=283) and a higher seroconversion rate (100%) than those of the baculovirus expressed full length Wisconsin HA (rHA, HA0 WI, Protein Sciences Corporation, Meriden, CT) at a 6 μg dose (GMT = 23, SC = 38%). Therefore, the D3 insertion format fusion protein that includes a portion of the globular head of influenza B Wisconsin (SEQ ID NO: 126) was more immunogenic in mice than the rHA WI (Protein Sciences) and was comparable to the inactivated B/HB virus antigen (obtained from CBER).

[00251] Multiple Victoria lineage strains were also evaluated. In particular, the Brisbane-like strains that were recommended for 2009-2011 trivalent influenza vaccines and are still recommended for a quadrivalent influenza vaccine in 2013-14 were evaluated. Having previously demonstrated improved potency of the B Brisbane-like strain B/Bangladesh/5945/2009 in mice, the immunogenicity of the Victoria prototypic fusion protein was evaluated in the rabbit model. Six NZW (New Zealand White) rabbits (3 of each gender) were treated i.m. with the indicated doses (6 μg, 12 μg, and 15 μg) of fusion proteins that include an antigen fused to a loop of domain 3 for either B/Brisbane/60/2008 (SEQ ID NO: 264) or

B/Bangladesh/5945/2009 (SEQ ID NO: 265). FLUVIRIN^® 201 1-12 (TIV, 15 μg total) was included as a positive immunogenicity control, and F147 buffer was included as a negative control. HAI titers were performed with ether-extracted virus matched to the composition: CBER anti B/Brisbane/60/2008 1 :5,120 or ferret anti B/Bangladesh/5945/2009 1 : 10,240. Data were shown as titers of individual rabbits with bars.

[00252] Horizontal lines represented geometric mean titers (GMT values above), and seroconversion percentage was provided in FIG. 23. As shown in FIG. 23, the Bangladesh fusion protein (HL787, SEQ ID NO: 158) was more immunogenic than the Brisbane fusion protein (HL657, SEQ ID NO: 128), with 100% seroconversion, when tested with the matched virus. The Brisbane-like Bangladesh strain, (D3Ins B/BR-like) as the Victoria lineage candidate may be useful in methods of providing protective immunity.

[00253] Fusion proteins of antigens fused to a loop of domain 3 were further evaluated for suitability across at least three recent and historical strains. Consistent with this strategy, testing of historical B candidates was carried out using D3Ins format of B/Sichuan/379/1999, a historical flu B virus (Yamagata lineage, STF2D3Ins B/SI, HL863, SEQ ID NO: 182). The portion of HA fused to the D3-ol site (see loop 6 of FIGs. 29 and 30) is SEQ ID NO: 56. HAI testing of mouse immune sera raised against B/Sichuan/379/1999 virus was performed. In this experiment, groups of 10 BALB/c mice were treated s.c. twice on days 0 and 21 with the STF2D3InsB/SI (HL903, SEQ ID NO: 183) fusion protein at a dose of 6 μg or with formulation buffer (F 147). Mice were bled on day 35. Serum HAI antibodies were measured using ether-extracted B/Sichuan/379/1999 virus, and values were plotted individually with GMTs and seroconversion indicated above. The immunogenicity results in FIG. 24 showed that the D3Ins composition of B Sichuan (SEQ ID NO: 182) at a 6 μg dose elicited a GMT of 132 with a

seroconversion rate of 100%. These data show that the D3Ins is suitable for use in generating compositions that include fusion protein of flagellin and portions of the globular head of influenza B for use in methods to treat seasonal exposure to influenza.

[00254] Example 7. Efficacy of fusion proteins with HA antigen inserted into loop regions of domain 3 of flagellin to shift pi to target influenza B in mice

[00255] In addition to immunogenicity testing of the D3Ins fusion proteins of the historical B virus, B/Sichuan/379/99 (D3Ins B/SI, HL863, SEQ ID NO: 182, Yamagata lineage) described above, efficacy testing in a lethal mouse challenge model was evaluated. Groups of 10 BALB/c mice were treated s.c. with the indicated doses (0.05 μg, 0.5 and 5 μg) of D3Ins B/SI 99 (HL863, SEQ ID NO: 182) fusion protein or F 147 buffer control on days 0 and 21, and bled on day 35. On day 42, mice were challenged intranasally with 5 x LD50 of B/Sichuan/379/1999 10⁵ pfu per mouse and monitored daily for mortality for 21 days and weight change for 14 days.

[00256] Survival rates and weights (mean percentage of initial weight) were plotted in FIGs. 25A-25B. Doses of about 5 μg and about 0.5 μg fusion protein D3Ins B/SI composition (two immunizations each) resulted in a 100% survival rate and about less than 12% weight loss in mice challenged with a lethal dose of B/Sichuan/379/99 virus (FIGs. 25A-25B). In contrast, the placebo group showed about a 20% survival rate and greater than or equal to about a 75% weight loss. These results indicated that the D3Ins fusion protein composition of

B/Sichuan/379/99 was efficacious at a submicrogram dose. The efficacy of D3Ins B/SI in this study thus supported the general suitability of D3Ins format for use in compositions that include portions of influenza B.

[00257] Example 8. Immunogenicity and safety of fusion proteins with HA antigen inserted into loop regions of domain 3 of flagellin to shift pi to target influenza B in a rabbit model

[00258] The immunogenicity of D3Ins fusion proteins was also evaluated in a rabbit model. A rabbit study designed to evaluate the immunogenicity of a range of doses of the R3 and D3Ins formats of B/WI was performed. HAI titers elicited by R3 and D3Ins formats of B Wisconsin in rabbits were measured. Fusion protein R3 B Wisconsin (HL724, SEQ ID NO: 152) and D3Ins B Wisconsin (HL772, SEQ ID NO: 126, a D3I-ol inserion, see Table 1) doses of 6, 12 and 18 μg were evaluated in groups of 6 New Zealand White rabbits, 3 of each gender (Covance Research Products, Denver, PA). For this study, fusion proteins were injected i.m. on days 0 and 21 , and serum was collected on day 35 for HAI analysis. A blend of baculovirus produced HA or egg-produced inactivated virus for matching strains (5 μg each) was included as a positive control, and F147 buffer was included as a negative control. HAI titers were performed with ether-extracted

B/Wisconsin/1/210 virus.

[00259] Data are shown as titers of individual rabbits, values plotted individually with GMTs and seroconversion indicated above (FIG. 26). One of the aims of the study was to confirm that the D3Ins fusion protein format of B/WI (HL772, SEQ ID NO: 126) was superior to the R3 format (HL724, SEQ ID NO: 152). As shown in FIG. 26, the fusion protein R3 format elicited low HAI titers and only achieved a high seroconversion rate (about 67%) at 18 μg with a GMT of about 22. By contrast, the fusion protein D3Ins format generated serconversion rates greater than or equal to about 50% at all three doses with a peak response at 12 μg (about 83%). This result confirmed previous mouse results demonstrating the superiority of the D3Ins format. The HL772 (SEQ ID NO: 126) may be useful in methods of treating subjects to provide protective immunity. [00260] In order to establish a suitable dose range for the rabbit model in terms of safety and immunogenicity, a dose range study with D3Ins B/WI (HL772, SEQ ID NO: 126) was performed. Groups of 10 NZW rabbits (5 of each gender) were treated i.m. with doses (3 μg, 6 μg, 9 μg, 12 μg, 15 μg, and 18 μg) of the

STF2D3Ins B/WI (HL772, SEQ ID NO: 126) fusion protein, inactivated B/Hubei- Wujiagang/158/2009 virus (5 μg Hubei virus, CBER), or formulation buffer (F147) on days 0 and 21, and bled on day 35. Serum HAI antibodies were measured by HAI test using ether-extracted B/Wisconsin/1/2010 virus, and plotted individually with GMTs and seroconversion.

[00261] The immunopotency results are shown in FIG. 27. Similar to the results with mice, strong HAI titers were elicited by doses in the low to mid microgram range, in this case with seroconversion of 100% occurring at doses as low as about 6 μg. The higher titers observed in this study were likely the result of a higher purity protein preparation. These results indicated that strong influenza B titers can be observed with low doses of the D3 insertion fusion protein that includes an influenza B antigen in multivalent mixes, thereby leading to a low total dose and a suitable safety window for multivalent blends for compositions that can be employed in methods of stimulating an immune response to the antigen, in particular providing protective immunity to a disease causing organism that includes the antigen.

[00262] The reactogenicity of D3Ins fusion proteins was examined in a rabbit safety model employing techniques previously been described (Taylor, D.N., et al, Vaccine 30:5761-5769 (2012)). In this model, three measures are found to be predictive of the appearance of adverse events in humans: poor appetite (low food consumption) within 24 hours of prime, increase in body temperature 6 hours after prime, and elevated serum CRP 24 hours post-prime.

[00263] Rabbits were treated with either formula buffer control (F147) as a negative control, or a fusion protein in which an antigen with an isoelectric point at least about 7.5 (B/Wisconsin/2010) is fused to a loop of domain 3 (HL772, SEQ ID NO: 126) at varying doses (3 μg, 6 μg, 9 μg, 12 μg, 15 μg, and 18 μg). Food consumption was measured about 24 hours after immunization (FIG. 28 A).

Temperature was measured rectally 6 hours post-immunization (FIG. 28B). CRP was measured from serum taken 24 hours after prime (FIG. 28C). Data for all measures were shown as results of individual rabbits with lines representing means and standard error of the mean. Dotted lines represent the safety threshold calculated using the data from formula control rabbits. Using the mean of the formulation control animals from several studies and using confidence intervals or multiples of standard deviations, safety thresholds were developed for each of the three measures. In this study a range of doses of fusion proteins in which antigens of influenza B of Wisconsin have been fused to loop 6 of domain 3 of flagellin (FIGs. 29 and 30) were evaluated

[00264] The results show that the highest doses of B Wisconsin were safe to a dose of about 18 μg (D3Ins B Wisconsin, HL772, SEQ ID NO: 126). Doses from about 3 to about 18 μg resulted in little decrease in food consumption (FIG. 28 A), with all group means falling within the safe threshold. Similarly, elevation in temperature at about 6 hours was minimal over the tested dose range (FIG. 28B), with all values below the safety thresholds. Finally serum CRP at about 24 hours showed only modest elevation up to about 18 μg (FIG. 28C). It is believed that doses as high as about 18 μg of D3Ins of B Wisconsin may be suitable (i.e., safe) as an influenza B component in a multivalent blend that can be employed in a method to stimulate an immune response to influenza B, in particular a protective immunity response to disease consequent or associated with influenza B infection.

[00265] The teachings of all patents, published applications and references cited herein are incorporated by reference in their entirety.

[00266] While this invention has been particularly shown and described with references to example embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the scope of the invention encompassed by the appended claims.

Claims

3710.10^4009

WO 2014/035989 PCT/US2013/056838

- 95 -

What is claimed is:

1. A composition that includes a fusion protein, comprising:

(a) a flagellin; and

(b) at least one antigen that has an isolectric point greater than about 7.0 and is fused to at least one loop of domain 3 of the flagellin, wherein the fusion protein activates a Toll-like Receptor 5.

2. The composition of Claim 1, wherein the isoelectric point of the antigen is at least one member selected from the group consisting of about 7.5, about 8.0, about 8.5, about 9.0, about 9.5 and about 10.0.

3. The composition of Claim 1, wherein the flagellin lacks at least one of a portion of a carboxy-domain 0 and portion of an amino-domain 0.

4. The composition of Claim 1 , further including an additional antigen fused to the flagellin at a site that is distinct from fusion of the antigen to the loop of domain 3 of the flagellin.

5. The composition of Claim 1, wherein the flagellin includes at least one

member selected from the group consisting of Salmonella typhimurium flagellin, an E coli flagellin, a S. muenchen flagellin, a Yersinia flagellin, a P. aeruginosa flagellin and a L. monocytogenes flagellin.

6. The composition of Claim 5, wherein the flagellin is the S. typhimurium

flagellin.

7. The composition of Claim 6, wherein the S. typhimurium flagellin is at least one member selected from the group consisting of SEQ ID NOs. 1 and 2.

8. The composition of Claim 7, wherein the S. typhimurium flagellin is SEQ ID NO: 2.

3710.10^400?

WO 2014/035989 PCT/US2013/056838

- 96 -

9. The composition of Claim 8, wherein the antigen to fused between amino acid residue 277 and amino acid residue 278 of SEQ ID NO: 2.

10. The composition of Claim 8, wherein the antigen to fused between amino acid residue 259 and amino acid residue 260 of SEQ ID NO: 2.

11. The composition of Claim 1, wherein the antigen is an influenza viral

antigen.

12. The composition of Claim 11 , wherein the influenza viral antigen is an

influenza B viral antigen.

13. The composition of Claim 11, wherein the influenza viral antigen is an

influenza A antigen subtype, and the influenza A antigen subtype is at least one member selected from the group consisting of H3, H7 and H9.

14. The composition of Claim 13, wherein the influenza A antigen subtype is H3.

15. The composition of Claim 1 1, wherein the influenza viral antigen is a

hemagglutinin antigen.

16. The composition of Claim 15, wherein the hemagglutinin antigen is at least one member selected from the group consisting of an influenza A viral hemagglutinin antigen and an influenza B viral hemagglutinin antigen.

17. The composition of Claim 15, wherein the hemagglutinin antigen is a portion of the hemagglutinin antigen that includes at least a portion of a globular head.

18. The composition of Claim 17, wherein the portion of the hemagglutinin lacks a transmembrane domain and a cytoplasmic domain.

19. The composition of Claim 18, wherein the portion of the hemagglutinin

further lacks an HA2 subunit.

3710.1054009.

WO 2014/035989 PCT/US2013/056838

- 97 -

20. The composition of Claim 19, wherein the portion of the globular head has at least one β-sheet at a bottom of the portion of the globular head.

21. The composition of Claim 20, wherein the portion of the globular head

further includes at least one β-sandwich at the bottom of the portion of the globular head.

22. The composition of Claim 21 , wherein the portion of the globular head

further includes at least two β-strands at the bottom of the portion of the globular head.

23. The composition of Claim 1 , further including an amino acid linker between at least one of an amino-terminus or a carboxy-terminus of the antigen and the loop of domain 3 of the flagellin.

24. A composition comprising at least three fusion proteins that each activate a Toll-like Receptor 5, wherein:

(a) a first fusion protein includes a first flagellin and a first influenza A viral hemagglutinin antigen that has an isoelectric point greater than about 6.0 fused to a portion of the first flagellin;

(b) a second fusion protein includes a second flagellin and a second

influenza A viral hemagglutinin antigen that is distinct from the first influenza A viral hemagglutinin antigen and that has an isoelectric point greater than about 7.0 fused to a portion of the second flagellin; and

(c) a third fusion protein includes a third flagellin and a first influenza B viral hemagglutinin antigen that has an isolectric point greater than about 8.0 and is fused to at least one loop of domain 3 of the third flagellin.

25. The composition of Claim 24, further including at least one adjuvant.

26. The composition of Claim 24, wherein:

3710.10^4ΟΩ?

WO 2014/035989 PCT/US2013/056838

- 98 -

(a) the first influenza A viral hemagglutinin antigen includes at least a portion or the entirety of an HA1 subunit that has at least a portion of a globular head that includes at least one β-sheet at a bottom of the globular head;

(b) the second influenza A viral hemagglutinin antigen includes at least a portion or the entirety of an HA1 subunit having at least a portion of a globular head that includes at least one β-sheet, at least one B- sandwich and at least two B-strands at the bottom of the portion of the globular head; and

(c) the first influenza B viral hemagglutinin antigen includes at least a portion or the entirety of an HA1 subunit that has at least a portion of a globular head that includes at least one β-sheet at a bottom of the globular head.

27. The composition of Claim 24, further including a fourth fusion protein that activates a Toll-like Receptor 5, the fourth fusion protein comprising a second influenza B viral hemagglutinin antigen that is distinct from the first influenza B viral hemagglutinin antigen and that has an isoelectric point greater than about 8.0 fused to at least one loop of domain 3 in the fourth flagellin.

28. The composition of Claim 27, wherein the second influenza B viral

hemagglutinin antigen includes at least a portion or the entirety of an HA1 subunit that has at least a portion of a globular head that includes at least one β-sheet at a bottom of the globular head.

29. A method of stimulating an immune response to an antigen in a subject, comprising the step of administering to the subject a composition that includes at least one fusion protein that activates Toll-like Receptor 5, the fusion protein comprising a flagellin and at least one antigen that has an isolectric point greater than about 7.0 and that is fused to at least one loop of domain 3 of the flagellin.

3710.1054002

WO 2014/035989 PCT/US2013/056838

- 99 -

30. The method of Claim 29, wherein the fusion protein of the composition

administered to the subject is in a dose of at least one member selected from the group consisting of about 1 μg, about 2 μg, about 3 μg, about 4 μg, about 5 μg, about 6 μg, about 7 μg, about 8 μg, about 9 μg, about 10 μg dose, about 15 μg dose, about 20 μg dose, about 25 μg dose, about 30 μg dose, about 35 μg dose, about 40 μg dose, about 45 μg dose and about 50 μg dose.

31. The method of Claim 29, wherein the composition is administered to the subject in a single dose.

32. The method of Claim 29, wherein the composition is administered to the subject in multiple doses.

33. The method of Claim 29, wherein the antigen is an influenza viral antigen.

34. The method of Claim 29, wherein the influenza viral antigen is at least a portion of hemagglutinin.

35. The method of Claim 29, wherein the composition includes an adjuvant.

36. A method of stimulating an immune response in a subject, comprising the step of administering to the subject a composition comprising at least three fusion proteins each that activate a Toll-like Receptor 5, wherein:

a) a first fusion protein includes a first flagellin and a first influenza A viral hemagglutinin antigen that has an isoelectric point greater than about 6.0 fused to a portion of the first flagellin;

(b) a second fusion protein includes a second flagellin and a second

influenza A viral hemagglutinin antigen that is distinct from the first influenza A viral hemagglutinin antigen and that has an isoelectric point greater than about 7.0 fused to a portion of the second flagellin; and

(c) a third fusion protein includes a third flagellin and a first influenza B viral hemagglutinin antigen that has an isolectric point greater than

3710.105400?

WO 2014/035989 PCT/US2013/056838

- 100 - about 8.0 and is fused to at least one loop of domain 3 of the third flagellin.

The method of Claim 36, wherein:

(a) the first influenza A viral hemagglutinin antigen administered to the subject includes at least a portion or the entirety of an HA1 subunit that has at least a portion of a globular head that includes at least one β-sheet at a bottom of the globular head;

(b) the second influenza A viral hemagglutinin antigen administered to the subject includes a at least a portion or the entirety of an HA1 subunit having at least a portion of a globular head that includes at least one β-sheet, at least one β-sandwich and at least two B-strands at the bottom of the portion of the globular head; and

(c) the first influenza B viral hemagglutinin antigen administered to the subject includes at least a portion or the entirety of an HA1 subunit that has at least a portion of a globular head that includes at least one β-sheet at a bottom of the globular head.

The method of Claim 36, wherein the composition administered to the subject further includes a fourth fusion protein that activates a Toll-like Receptor 5, the fourth fusion protein comprising a second influenza B viral hemagglutinin antigen that is distinct from the first influenza B viral hemagglutinin antigen and that has an isoelectric point greater than about 8.0 fused to at least one loop of domain 3 in the fourth flagellin.

The method of Claim 38, wherein the second influenza B viral

hemagglutinin antigen administered to the subject includes at least a portion or the entirety of an HA1 subunit that has at least a portion of a globular head that includes at least one β-sheet at a bottom of the globular head.